Industrial simulation method for mechanical structure

Abstract

The invention relates to the technical field of mechanical engineering and computer-aided engineering simulation, and discloses a mechanical structure industrial simulation method which comprises the following steps: S1, establishing a geometric model, material attributes and boundary conditions of a mechanical structure; s2, constructing a nonlinear multi-physics field coupling model; s3, constructing a micro-scale model, and describing a grain slippage behavior and a phase change characteristic in the material and the influence of the grain slippage behavior and the phase change characteristic on a macroscopic multi-physical field; s4, mapping the microscale model to a macroscopic field through a homogenization method, and establishing a microcosmic-to-macroscopic multi-field coupling model; s5, dynamically solving the coupling model, and optimizing the solving process by adjusting the time step length and the boundary condition in real time; and S6, monitoring the dynamic response of the multiple physical fields in real time based on stability analysis. According to the method, nonlinear coupling analysis among a heat conduction field, a mechanical field and a vibration field is realized, and mutual influence and dynamic response among multiple fields under complex working conditions are effectively captured through the method.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical engineering and computer-aided engineering simulation, and in particular to a mechanical structure industrial simulation method. Background Art

[0002] In modern industry, simulation technology of mechanical structures is widely used in many scenarios. For example, aircraft engine components need to withstand complex thermal-mechanical coupling loads in high temperature and high pressure environments; in automobile structural design, body parts need to consider collision mechanics analysis and dynamic simulation of vibration response; energy equipment such as gas turbines or nuclear reactors also require detailed evaluation of the multi-field coupling effects of their structures under extreme working conditions. In addition, in industrial manufacturing, fatigue life prediction and performance optimization of high-precision mechanical parts are also highly dependent on multi-physics simulation tools. These scenarios require simulation technology to accurately simulate the interaction of physical fields such as heat conduction, mechanical stress, and vibration, and to carefully model multi-scale characteristics.

[0003] However, in the existing technology, for the above-mentioned industrial application scenarios, most simulation tools only support the analysis of a single physical field, such as separate heat conduction analysis or mechanical stress analysis of mechanical parts. This type of method has been widely used under simple working conditions, such as the simulation of the thermal conductivity performance of fuel rods in nuclear reactors. However, when it comes to complex multi-field interactions, such as the thermal expansion and stress coupling effects of aircraft engine blades, this single physical field simulation can no longer meet engineering needs. Since the dynamic effects between different physical fields are ignored, traditional solutions have difficulty capturing key performance details, resulting in low accuracy of simulation results. Summary of the invention

[0004] In order to make up for the above deficiencies, the present invention provides a mechanical structure industrial simulation method, which aims to improve the problems in the prior art such as insufficient accuracy of multi-physical field coupling analysis of mechanical structures, the inability to effectively map the microscopic properties of materials to macroscopic responses, and low solution stability and efficiency under dynamic conditions.

[0005] In a first aspect, the present invention provides the following technical solution, a mechanical structure industrial simulation method, comprising the following steps:

[0006] S1: Establish the geometric model, material properties and boundary conditions of the mechanical structure;

[0007] S2: Construct a nonlinear multi-physics field coupling model, including heat conduction field, mechanical field and vibration field. The heat conduction field is modeled based on the thermal conductivity of the material, the thermal effect caused by mechanical work and the external heat source. The mechanical field is modeled based on the nonlinear elastic-plastic relationship and thermal expansion effect. The vibration field is modeled based on the temperature-related stiffness matrix and external dynamic load.

[0008] S3: Construct a micro-scale model to describe the internal grain sliding behavior, phase change characteristics of the material and its impact on macroscopic multi-physics fields;

[0009] S4: Through the homogenization method, the micro-scale model is mapped to the macroscopic field, and a multi-field coupling model from micro to macro is established;

[0010] S5: Dynamically solve the coupled model and optimize the solution process by adjusting the time step and boundary conditions in real time;

[0011] S6: Real-time monitoring of dynamic responses of multi-physics fields based on stability analysis.

[0012] Preferably, in step S2, the heat conduction field is modeled by combining material density, specific heat capacity, thermal conductivity with mechanical work and external heat source, and boundary conditions include the relationship between normal heat flux density and ambient temperature and convection heat transfer coefficient.

[0013] Preferably, in step S2, the mechanical field is constructed based on a nonlinear elastic-plastic relationship, which includes the initial yield strength, strain hardening coefficient, strain rate sensitivity and thermal softening effect of the material, and also includes the thermal expansion stress caused by the thermal expansion coefficient and elastic modulus.

[0014] Preferably, in step S2, the mechanical field is constructed based on a nonlinear elastic-plastic relationship, which includes the initial yield strength, strain hardening coefficient, strain rate sensitivity and thermal softening effect of the material, and also includes the thermal expansion stress caused by the thermal expansion coefficient and elastic modulus.

[0015] Preferably, in step S3, the micro-scale model is modeled based on material grain size, slip resistance and thermal strengthening effect, and is further optimized by describing the change process of phase transformation volume fraction.

[0016] Preferably, in step S4, the mapping between the micro-scale model and the macroscopic multi-physical field is achieved by a homogenization method, and the homogenization method is used to convert the microscopic stress distribution into a macroscopic stress component to ensure the continuity of the microscopic effect in the macroscopic field.

[0017] Preferably, in step S5, the dynamic solution is combined with a deep learning model, and a neural network is used to accelerate the mapping process from the micro-scale model to the macro-scale model, and to achieve real-time response under complex working conditions.

[0018] In a second aspect, the present invention provides the following technical solution, a mechanical structure industrial simulation system, the system comprising:

[0019] An input module for receiving a geometric model, material properties and boundary conditions of a mechanical structure;

[0020] A modeling module, used to construct a nonlinear multi-physics field coupling model, wherein the model includes a heat conduction field, a mechanical field, and a vibration field;

[0021] Microscopic modeling module, used to describe the material grain slip behavior, phase change characteristics, and establish microscopic scale models;

[0022] A mapping module for mapping micro-scale models to macroscopic multiphysics through homogenization methods;

[0023] Solving module, used for dynamic solving of nonlinear multi-physics coupling models;

[0024] Optimization module for adjusting simulation parameters based on real-time feedback, including dynamic optimization of time steps, boundary conditions, and stiffness matrices.

[0025] In a third aspect, the invention provides the following technical solution: a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned mechanical structure industrial simulation method when executing the computer program.

[0026] In a fourth aspect, the present invention provides the following technical solution: a readable storage medium having a computer program stored thereon, wherein the computer program implements the above-mentioned mechanical structure industrial simulation method when executed by a processor.

[0027] The present invention has the following beneficial effects:

[0028] 1. The present invention realizes the nonlinear coupling analysis between the heat conduction field, the mechanical field and the vibration field. Through this method, the mutual influence and dynamic response between multiple fields under complex working conditions are effectively captured. Compared with the technical solutions in the prior art that only deal with a single physical field or weak coupling analysis, the problem of ignoring the correlation between physical fields is avoided, and the problem of low overall simulation accuracy is solved.

[0029] 2. The present invention accurately characterizes the influence of the microscopic characteristics of the material on the overall structural response, especially in terms of grain slip behavior and dynamic phase change characteristics, through the homogenization mapping of the micro-scale model and the macro-model. Compared with the technical solution of the prior art that directly uses the macro-model and ignores the microscopic characteristics, the present invention significantly improves the simulation's ability to capture the dynamic effects of local material strengthening and phase change, and solves the limitation that the existing method is difficult to reflect the microscopic influence.

[0030] 3. In the present invention, Lyapunov stability analysis and machine learning algorithm are combined in the dynamic solution process to timely adjust the time step and system parameters to ensure the convergence and stability of the simulation solution under complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A method flow chart of a mechanical structure industrial simulation method proposed by the present invention;

[0032] Figure 2 This is a schematic diagram of the system framework of a mechanical structure industrial simulation system proposed by the present invention. DETAILED DESCRIPTION

[0033] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] Embodiment 1

[0035] Reference Figure 1 In a first embodiment of the present invention, the present invention provides a mechanical structure industrial simulation method, comprising the following steps:

[0036] S1: Establish the geometric model, material properties and boundary conditions of the mechanical structure;

[0037] Specifically, in step S1, when establishing the geometric model of the mechanical structure, it is necessary to ensure the topological integrity of the model and refine the local structure to meet the needs of multi-scale analysis. In this step, the mechanical parts are three-dimensionally modeled by geometric modeling tools or CAE software, and the complex geometric shapes are reasonably simplified according to the simulation requirements, but the accurate expression of key parts must still be guaranteed. For example, the tooth shape in the gear meshing area needs to be modeled with high precision so that the force field distribution and temperature rise changes can be accurately captured later.

[0038] In this embodiment, the geometric model can adopt a standard parametric modeling method to input the dimensions, structural characteristics and functional design parameters of the mechanical parts. These parameters include but are not limited to length, width, height, wall thickness, bending radius, aperture, groove shape, etc. In some embodiments, reverse engineering technology can be combined, for example, the geometric shape data of existing mechanical parts can be obtained through a three-dimensional scanning device, and a three-dimensional model file can be generated in the CAD software. As an option, for complex mechanical parts (such as multi-stage gear systems or multi-link mechanisms), it can be further decomposed into multiple substructure models, and the connection relationship between components can be defined by assembly constraints.

[0039] The setting of material properties is another key link in simulation modeling. In this embodiment, material properties need to be specified for each mechanical component. Specifically, these material properties include but are not limited to density ρ (in kg / m 3), elastic modulus E (unit: Pa), Poisson's ratio ν (dimensionless), thermal expansion coefficient α (unit: 1 / K), thermal conductivity k (unit: ), specific heat capacity c p (Unit is ), yield strength A (in Pa), etc. In a possible implementation, dynamic properties can be assigned to different materials, such as the temperature-related elastic modulus E(T), and the formula can be:

[0040] E(T)=E0·(1-βT)

[0041] Among them, E0 is the initial elastic modulus of the material (in Pa), β is the temperature attenuation coefficient (dimensionless), and T is the temperature (in K). This setting method can support the analysis requirements of multi-physics field coupling in subsequent steps.

[0042] In some embodiments, multi-layer material properties can also be set for the composite material, and the properties of each layer of material can be defined independently. For example, the thermal conductivity and elastic modulus between composite material layers can be assigned different values ​​to simulate the difference in thermal resistance and mechanical response between layers.

[0043] Regarding the setting of boundary conditions, in general, reasonable boundary constraints should be set according to the force, temperature, and vibration environment of the mechanical components in actual operating conditions. In this embodiment, the boundary conditions include two types: mechanical boundary conditions and thermal boundary conditions. The mechanical boundary conditions can be set as fixed constraints, sliding constraints, or rotational constraints. For example, when simulating gear meshing, the gear bearing seat can be subject to fixed constraints, and the gear meshing surface can be defined as a contact boundary condition.

[0044] In addition, dynamically changing load curves can be taken into account when applying loads, such as periodic loads according to a sinusoidal function:

[0045] F(t)=F0·sin(ωt)

[0046] Where F0 is the load amplitude (in N), ω is the angular frequency (in rad / s), and t is the time (in s). The thermal boundary conditions mainly include the setting of heat flux density and convection heat transfer coefficient. In a possible implementation, the setting of turbulence density adopts the following formula:

[0047] q=h·(T s -T ∞ )

[0048] Where q is the heat flux (in W / m 2 ), h is the convective heat transfer coefficient (unit: W / m 2 ·K),T s is the surface temperature of the mechanical structure (in K), T∞ is the ambient temperature (in K).

[0049] In some embodiments, the thermal boundary conditions can be set to dynamic distribution in combination with sensor data to simulate the thermal load changes of mechanical components under high temperature and high pressure conditions.

[0050] As an option, boundary friction conditions can also be defined for specific mechanical components (such as high-speed rotating machinery). Specifically, the friction coefficient μ of the gear meshing surface can be expressed by the following formula:

[0051] μ=μ0·(1-γT)

[0052] Wherein, μ0 is the initial friction coefficient (dimensionless), and γ is the temperature attenuation coefficient of the friction coefficient (dimensionless). In some embodiments, multiple boundary conditions can be set for mechanical components. For example, the gear meshing surface with both thermal boundary conditions and mechanical boundary conditions can consider both the thermal brittleness and contact stress distribution. In this case, the thermal-mechanical boundary conditions need to be uniformly expressed through a multi-physics field coupling model to support the accuracy of subsequent modeling.

[0053] S2: Construct a nonlinear multi-physics field coupling model, including heat conduction field, mechanical field and vibration field. The heat conduction field is modeled based on the thermal conductivity of the material, the thermal effect caused by mechanical work and the external heat source. The mechanical field is modeled based on the nonlinear elastic-plastic relationship and thermal expansion effect. The vibration field is modeled based on the temperature-related stiffness matrix and external dynamic load.

[0054] Specifically, after completing the setting of the geometric model, material properties and boundary conditions in step S1, it is necessary to further perform nonlinear multi-physics field coupling modeling on the thermal conduction field, mechanical field and vibration field of the mechanical structure. Generally, multiple physical fields do not exist independently, but are strongly coupled through a variety of physical effects. For example, mechanical work will cause temperature to rise, and temperature changes will cause material expansion, which will in turn affect vibration characteristics and structural stress distribution.

[0055] In this embodiment, a heat conduction field model is first established to describe the temperature field distribution of mechanical components during operation. The heat conduction field is modeled based on the classic Fourier law, and its control equation is:

[0056]

[0057] Where T represents the temperature field, the unit is K;

[0058] ρ represents the material density, in kg / m 3 ;

[0059] c pIndicates the specific heat capacity of the material, in J / (kg·K);

[0060] k is thermal conductivity, unit is W / (m·K);

[0061] σ: represents the heat converted from mechanical work, where σ is the stress tensor, is the strain rate tensor;

[0062] Q represents the external heat source density, in W / m 3 .

[0063] Alternatively, the heat flux can be specified using boundary conditions, such as the convection heat transfer boundary condition:

[0064]

[0065] Where h is the convective heat transfer coefficient, in units of

[0066] T ∞ is the ambient temperature, in K;

[0067] is the normal heat flux, in W / m 2 .

[0068] In modeling the thermal conductivity field, some embodiments may also consider the temperature-dependent properties of the material. For example, thermal conductivity k and specific heat capacity c p It can be set as a function of temperature to more accurately describe heat transfer behavior in high temperature environments.

[0069] Specifically, the construction of the mechanical field is based on nonlinear elastic-plastic theory. The mechanical field is described by the Cauchy stress equilibrium equation, whose mathematical expression is:

[0070]

[0071] Where, σ represents the stress tensor, in Pa;

[0072] f represents body density, unit is N / m 3 ;

[0073] u represents the displacement field, and its unit is m.

[0074] In this embodiment, the stress tensor is calculated by a nonlinear constitutive relationship using the JohnsonCook model:

[0075]

[0076] Where A is the initial yield strength of the material, in Pa;

[0077] B is the material strengthening coefficient, in Pa;

[0078] n is the strain hardening exponent, dimensionless;

[0079] C is the strain rate sensitivity coefficient, dimensionless;

[0080] ∈ is the dimensionless equivalent plastic strain; is the equivalent plastic strain rate, in s - 1; T is the current temperature, in K; T m is the melting temperature of the material, in K; m is the thermal softening index, dimensionless.

[0081] In one possible implementation, the mechanical field also includes the influence of thermal expansion stress. Thermal expansion stress can be expressed by the following formula:

[0082] σ T =αE(T-T0)

[0083] Where α is the thermal expansion coefficient, in units of E represents the elastic modulus, the unit is Pa; T0 represents the initial temperature, the unit is K.

[0084] The modeling of the vibration field is used to describe the vibration response of mechanical parts under dynamic loads. In this embodiment, the dynamic equation of the vibration field is:

[0085]

[0086] Among them, M represents the mass matrix, the unit is kg; C represents the damping matrix, the unit is kg / s; K(T) represents the stiffness matrix, its size is related to the temperature T; F(t) is the dynamic load, the unit is N.

[0087] In some embodiments, in order to achieve strong coupling of the three fields of heat, force and vibration, the temperature change of the thermal field is linked with the mechanical field and the vibration field. For example, the high temperature area in the temperature field will cause the material stiffness to decrease, thereby affecting the amplitude of the vibration response. In addition, the stress change in the force field will induce a heat source effect, which directly affects the distribution of the thermal field.

[0088] As an option, the model parameters can be further corrected by an optimization algorithm. For example, when the dynamic loads vary in a complex manner, the initial and boundary conditions in the model can be adjusted to ensure the convergence of the coupled field solution.

[0089] S3: Construct a micro-scale model to describe the internal grain sliding behavior, phase change characteristics of the material and its impact on macroscopic multi-physics fields;

[0090] Specifically, on the basis of completing the multi-physics field coupling model in step S2, step S3 aims to further refine the modeling accuracy and capture the physical behavior of the internal materials of the mechanical structure at the microstructural level such as grains and phase changes by establishing a microscale model. In general, the macroscopic response of mechanical components is not only affected by external boundary conditions, but also significantly restricted by the internal microstructure of the material (such as grain slip and phase change behavior). Therefore, this step provides a physical basis for the refined analysis of multi-physics fields through the construction of a microscale model.

[0091] In this embodiment, the core of the micro-scale model is to describe the grain sliding behavior of the material and its evolution law under different temperature and stress conditions.

[0092] Specifically, the slip resistance can be calculated by the following formula:

[0093] τ=τ0+k·d -1 / 2 +α·ΔT

[0094] in:

[0095] τ is the sliding resistance, in Pa;

[0096] τ0 is the initial sliding resistance, in Pa;

[0097] k is the grain strengthening coefficient, unit is Pa·m 1 / 2;

[0098] d is the grain size, in m;

[0099] α is the thermal enhancement coefficient, in Pa / K;

[0100] ΔT=T-T0, where T is the current temperature in K, and T0 is the initial reference temperature in K.

[0101] As an option, the distribution of grain size d can be further set as a random variable, and the distribution of grains of different sizes can be described by a probability density function. In some embodiments, commonly used distribution forms include normal distribution and log-normal distribution, which are closer to the microstructural characteristics of actual materials. Specifically, materials undergo phase change behavior under high temperature or high stress conditions, and this phase change manifests itself as the formation of a new phase and the evolution of its volume fraction at a microscopic scale. In this embodiment, the dynamic change of the phase change volume fraction f can be described by the following formula.

[0102]

[0103] in:

[0104] f is the phase change volume fraction (dimensionless, ranging from 0 to 1);

[0105] t is time, in seconds;

[0106] L is the phase change rate constant, in s -1 ·K -1 ;

[0107] T is the current temperature, in K;

[0108] T c is the critical temperature of phase transition, in K.

[0109] In a possible implementation, different phase change rate constants L and critical temperatures T can be set according to the specific material type. c , to adapt to different microscopic phase transformation characteristics such as martensitic transformation or austenitic transformation. In some embodiments, the energy change of the microstructure during the phase transformation process can be more accurately described by introducing an activation energy parameter.

[0110] In order to combine the micro-scale with the macro-response, this embodiment further models the contribution of grain slip behavior and phase change to the macro-properties of the material. Specifically, the yield strength of the material can be expressed as the sum of the slip resistance and the phase change factor:

[0111] σ y =τ+β·f

[0112] in:

[0113] σ y is the yield strength, unit is Par;

[0114] β is the phase change enhancement factor, in Pa;

[0115] As an option, the slip resistance and phase change factor can be defined separately for the multilayer structure of the composite material, and the change in its overall yield strength can be described by the interlayer coupling formula. For example, assuming that the multilayer structure of the composite material consists of three layers, the yield strength can be calculated by the weighted formula:

[0116]

[0117] in:

[0118] represents the yield strength of the i-th layer, in Pa;

[0119] w i represents the weight factor of the i-th layer (dimensionless, and satisfies w1+w2+w3=1).

[0120] In the numerical implementation of the micro-scale model, the finite element method can be combined with the present embodiment for discretization calculation. In some embodiments, for complex grain structures, a grain partitioning method based on the Voronoi diagram can be used to simulate the real geometric shape of polycrystalline materials.

[0121] The elements generated by the Vorono i diagram can be directly applied to the finite element solution of the thermal and mechanical fields, thereby accurately coupling the thermal-mechanical behavior at the grain level to the overall analysis.

[0122] Specifically, in order to further improve the solution efficiency, the reduced-order modeling technology can be used in this embodiment to reduce the computational complexity of the micro-scale model. As a possible implementation method, the slip behavior and phase change process can be accelerated by combining a data-driven method with a machine learning model (such as a convolutional neural network), thereby significantly reducing the computational time in large-scale multi-physics field simulations.

[0123] S4: Through the homogenization method, the micro-scale model is mapped to the macroscopic field, and a multi-field coupling model from micro to macro is established;

[0124] Specifically, after completing the construction of the micro-scale model in step S3, it is necessary to map the characteristics of the micro-model to the macroscopic physical field to achieve an organic connection between the microstructural characteristics and the macroscopic multi-physical field coupling. In general, the micro-scale model includes refined behaviors such as grain slip and phase change, but directly using it for large-scale macroscopic calculations will significantly increase the computational complexity. Therefore, using the homogenization method to process the micro-scale model can effectively simplify the calculation while ensuring that the physical influence of the microscopic characteristics on the macroscopic response is accurately transmitted.

[0125] In this embodiment, the core of the homogenization method is to calculate the equivalent properties of the material in the macroscopic field through the mathematical integration of micromechanics, thermal and phase change behaviors. Specifically, the transition from microscopic to macroscopic fields can be achieved through representative volume units. The selection of RVE should be small enough to capture microscopic properties and large enough to be statistically representative.

[0126] In terms of mechanical fields, the stress distribution of microscopic grains can be used to calculate the macroscopic stress tensor <σ> through the homogenization method.

[0127] Specifically, the calculation formula of the macroscopic stress tensor is as follows:

[0128]

[0129] in:

[0130] <σ> is the macroscopic stress tensor, in Pa;

[0131] V is the volume of the representative volume unit, in m3 ;

[0132] σ m is the micro stress component, unit is Pa;

[0133] dV is the integrated volume unit, in m 3 .

[0134] As an option, the microscopic stress field in the RVE can be numerically discretized, for example, by using the finite element method (FEM) to calculate the stress distribution of each unit, and then overall homogenization can be achieved by summing. In some embodiments, the characteristics of different grains can be simulated by random distribution to improve the accuracy of the homogenization result in view of the differences in grain orientation in polycrystalline materials.

[0135] Specifically, for the thermal field, the thermal conductivity k in the microscale model is m The equivalent thermal conductivity k can be calculated by the homogenization method eff .

[0136] In this embodiment, the calculation formula of the equivalent thermal conductivity is as follows:

[0137]

[0138] in:

[0139] k eff is the equivalent thermal conductivity, unit is W / (m·K);

[0140] k m It is the thermal conductivity of the microscopic unit, and its unit is W / (m·K).

[0141] In one possible implementation, the equivalent thermal conductivity can be derived by calculating the temperature gradient and heat flux density of the RVE. For example, suppose there is a temperature gradient inside the RVE By calculating the heat flux Combined with the volume weighted average, we finally get k eff .

[0142] For the phase change characteristics of materials, the phase change volume fraction f in the microscale model m It also needs to be mapped to the macroscopic field. In this embodiment, the equivalent phase change volume fraction f eff It can be calculated by the following formula:

[0143]

[0144] in:

[0145] f eff is the equivalent phase change volume fraction (dimensionless);

[0146] f m is the volume fraction of microscopic phase change (dimensionless).

[0147] As an option, a dynamic phase change rate model can be further combined to update the equivalent phase change volume fraction in real time in the macroscopic field. For example, for rapid heating or cooling processes, the phase change rate constant L and critical temperature Tc can be corrected in combination with the actual temperature distribution, thereby improving the mapping accuracy of the phase change behavior.

[0148] In this embodiment, in order to achieve more efficient micro-macro field mapping, a data-driven machine learning method can be used. For example, a convolutional neural network is used to train micromechanics and thermal behavior to quickly predict equivalent parameters. This method can not only significantly reduce the computational cost, but also maintain a high prediction accuracy under multiple working conditions.

[0149] In some embodiments, a hybrid multi-scale modeling method can be combined to use the homogenization results as input parameters of the macro field model, and further refine the model to correct the abnormal behavior of the local area. For example, in the mechanical contact area, due to the significant stress concentration and frictional heat effect, the micro details can be locally restored in the macro field model to ensure the accuracy of the overall simulation.

[0150] Specifically, in order to verify the accuracy of the homogenization results, the calculation results of the macroscopic field can be compared with the microscopic field. For example, when simulating high-temperature load conditions, the consistency of the grain slip stress and the overall yield strength can be compared to adjust the parameter settings of the homogenization method.

[0151] S5: Dynamically solve the coupled model and optimize the solution process by adjusting the time step and boundary conditions in real time;

[0152] Specifically, after completing the homogenization mapping from the microscale to the macroscopic field in step S4, the coupling model needs to be dynamically solved to obtain the time evolution process of the mechanical structure under the coupling of multiple physical fields. In general, the solution of the multi-physical field model needs to simultaneously deal with the nonlinear relationship between the thermal field, the mechanical field, and the vibration field, while considering the dynamic adjustment of the time step and the real-time optimization of the boundary conditions to ensure the convergence and stability of the solution. The implementation of this step plays a key role in obtaining high-precision simulation results.

[0153] In this embodiment, the dynamic solution uses the finite element method to discretize the space and combines the implicit time integration method to solve the time. Specifically, the dynamic control equation of the thermal-mechanical-vibration coupling model is:

[0154]

[0155] in:

[0156] M(T) is the mass matrix, whose size is dynamically adjusted with the change of temperature T, and the unit is kg;

[0157] C is the damping matrix, in kg / s;

[0158] K(T) is the stiffness matrix, and its size is also dynamically adjusted with the temperature T, and its unit is N / mas;

[0159] u is the displacement vector, in m;

[0160] F(t) is the time-dependent external load vector, in N.

[0161] Specifically, the temperature dependence of the stiffness matrix K(T) can be defined by the following formula:

[0162] K(T)=K0(1-βT)

[0163] in:

[0164] K0 is the initial stiffness matrix, in N / mi;

[0165] β is the temperature stiffness reduction coefficient, dimensionless;

[0166] T is the current temperature in K.

[0167] As an option, the time step in the dynamic solution can be adjusted in real time according to the system response characteristics. In some embodiments, the time step Δt can be dynamically adjusted according to the following conditions:

[0168]

[0169] in:

[0170] F(t)=F0sin(ωt)

[0171] Δt is the current time step, in seconds;

[0172] ∈ is the error tolerance, dimensionless;

[0173] is the modulus of the acceleration vector, in m / s 2 ;

[0174] Δt max is the maximum allowed time step, in seconds.

[0175] Specifically, the external load F(t) can be set as a dynamically changing periodic load. For example, in vibration analysis, a periodic load in the form of a sinusoidal function can be used:

[0176] F(t)=F0sin(ωt)

[0177] in:

[0178] F0 is the load amplitude vector, in N;

[0179] ω is the angular frequency, in rad / s;

[0180] t is time, unit is s.

[0181] In one possible implementation, the solution process can be further combined with boundary condition optimization to improve the calculation accuracy and convergence. For example, for the thermal boundary conditions on the surface of a mechanical structure, the heat transfer coefficient h can be dynamically adjusted according to the actual temperature field distribution. This dynamic adjustment can be achieved through the following relationship:

[0182] h=h0(1+γT)

[0183] in:

[0184] h is the heat transfer coefficient after real-time adjustment, in W / (m 2 K);

[0185] h0 is the initial heat transfer coefficient, unit is W / (m 2 K);

[0186] γ is the temperature sensitivity of the heat transfer coefficient and is dimensionless.

[0187] In some embodiments, in order to accelerate the solution of complex multi-physics fields, reduced-order modeling technology can be combined. Specifically, some links of dynamic solution can be accelerated through a prediction model based on deep learning, such as using a neural network to predict the distribution of the temperature field and reduce the number of solutions to the heat conduction equation. In this embodiment, the input of the neural network can include geometric shapes, material properties and initial conditions, and the output is the corresponding temperature distribution field. In order to further ensure the stability of the solution, Lyapunov stability analysis is also combined in this embodiment to judge the stability of the system by real-time monitoring of the derivative of the Lyapunov function.

[0188] In this embodiment, the value of the Lyapunov function is calculated in real time at each time step of the solution, and the boundary conditions or time step are adjusted according to the derivative trend. If a positive value appears, immediately shorten the time step or modify the stiffness matrix to avoid system divergence.

[0189] S6: Monitor the dynamic response of multi-physics fields in real time based on stability analysis, optimize system parameters through feedback, and ensure simulation stability.

[0190] Specifically, after completing the dynamic solution of the multi-physics coupling model in step S5, in order to ensure the accuracy and stability of the simulation process, it is necessary to monitor and adjust the dynamic response during the solution process in real time. In general, the coupling model of multiple physical fields may have problems such as numerical instability and boundary condition offset under complex working conditions. If these problems are not optimized in time, they may lead to a decrease in the accuracy of the solution or complete divergence. Therefore, real-time monitoring of the stability of the system and optimizing the simulation parameters in combination with the feedback mechanism are the key to achieving overall simulation reliability.

[0191] In this embodiment, in order to monitor the dynamic response of the system, Lyapunov stability analysis is used as the core method. Lyapunov stability theory can judge the stability of the system by constructing a criterion in the form of an energy function. Specifically, the Lyapunov function V is defined as:

[0192]

[0193] in:

[0194] V is the Lyapunov function, dimensionless;

[0195] u is the displacement vector, in m;

[0196] is the velocity vector, in m / s;

[0197] M is the mass matrix, in kg;

[0198] κ is the thermal field stability coefficient, dimensionless;

[0199] T is the current temperature, in K;

[0200] In general, when the time derivative of the Lyapunov function satisfies the following conditions:

[0201]

[0202] The system can be considered stable in the current time step.

[0203] In this embodiment, during the solution process of each time step, the value of the Lyapunov function and its derivative are calculated in real time. A positive value indicates that the system may be unstable. At this time, the key parameters are adjusted through the feedback optimization mechanism. As a possible adjustment method, the time step Δt can be optimized through the following strategy:

[0204]

[0205] in:

[0206] Δt new is the adjusted time step, in seconds;

[0207] Δt current is the current time step, in seconds;

[0208] α is the step size adjustment coefficient, dimensionless;

[0209] ∈ is a small positive value that prevents the denominator from being zero and is dimensionless.

[0210] As an option, the state of the system can also be adjusted in combination with real-time optimization of boundary conditions. For example, for the constraints in the mechanical field, when an oscillation trend is detected in the boundary constraints, the oscillation can be eliminated by increasing the damping coefficient C. The adjustment formula is as follows:

[0211]

[0212] in:

[0213] C new is the adjusted damping coefficient, in kg / s;

[0214] C current is the current damping coefficient, in kg / s;

[0215] γ is the damping adjustment coefficient, dimensionless.

[0216] As a further implementation method, the feedback optimization mechanism can be extended to a multi-scale solution framework. Specifically, by introducing adaptive control parameters between the microscopic field and the macroscopic field, such as real-time adjustment of the microscopic phase change rate and the macroscopic yield strength, the flexibility and stability of the multi-physics field solution are further enhanced. In some embodiments, the results of the feedback optimization can also be subsequently verified. For example, after completing each round of feedback adjustment, the effect of parameter optimization can be evaluated by comparing the difference between the simulation results and the experimental data. If it is detected that the error exceeds the preset range, the optimization strategy can be further refined or more feedback paths can be added.

[0217] Embodiment 2:

[0218] Reference Figure 2 In a second embodiment of the present invention, the present invention provides a mechanical structure industrial simulation system, the system includes an input module, a modeling module, a micro-modeling module, a mapping module, a solving module, and an optimization module.

[0219] The input module is the starting point of the system, which receives the geometry model, material properties and boundary conditions of the mechanical structure. The geometry model can be imported in a variety of formats and is pre-processed to ensure integrity. Material properties include density, elastic modulus, thermal conductivity, etc. Boundary conditions are used to describe the working conditions, such as loads, constraints and thermal boundaries. The input module passes this data to the modeling module to provide a basis for subsequent construction.

[0220] The modeling module builds a nonlinear multi-physics field coupling model, including heat conduction field, mechanical field and vibration field. The heat conduction field is used to simulate temperature distribution, the mechanical field describes stress, strain and thermal expansion effects, and the vibration field is used for dynamic response analysis. The coupling between modules ensures the mutual influence between physical fields, such as the effect of thermal expansion on stress. The modeling results are provided to the microscopic modeling module for further refined description.

[0221] The microscopic modeling module describes the grain slip behavior and phase change characteristics of the material. These microscopic characteristics are constructed through independent models, such as grain slip and phase change volume fraction. The module outputs microscopic results for homogenization processing to convert microscopic characteristics into macroscopic equivalent parameters.

[0222] The mapping module maps the microscopic model results to the macroscopic multi-physics field through the homogenization method. The microscopic parameters such as stress, thermal conductivity and phase change volume fraction are equivalently calculated by RVE (representative volume element) to ensure that the macroscopic field reflects the microscopic characteristics. The mapping results are passed to the solution module to achieve the unification of multi-scale models.

[0223] The solution module receives the mapped macro model and dynamically solves the coupling field. Numerical methods such as the finite element method are used to solve the spatiotemporal evolution of the temperature field, stress field and dynamic response. The coupling field state is updated in real time during the solution process, and the simulation parameters are adjusted in collaboration with the optimization module.

[0224] The optimization module adjusts the time step, boundary conditions and model parameters by real-time monitoring of dynamic responses, such as temperature field, displacement field or modal vibration. Parameter adjustment is carried out according to the feedback closed-loop optimization mechanism to ensure the stability and accuracy of the solution process. The optimization results directly act on the solution module to improve overall efficiency and reliability.

[0225] Embodiment 3

[0226] The third embodiment of the present invention is based on the same inventive concept and proposes a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of a mechanical structure industrial simulation method of the above embodiment are implemented.

[0227] Embodiment 4

[0228] The fourth embodiment of the present invention is based on the same inventive concept. The present invention proposes a computer device, which includes: a processor and a memory; the processor and the memory communicate with each other; the memory is used to store instructions; the processor is used to execute the instructions in the memory, and execute a mechanical structure industrial simulation method of the above embodiment.

[0229] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0230] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A mechanical structure industrial simulation method, characterized in that: The following steps are involved: S1: Establish the geometric model, material properties and boundary conditions of the mechanical structure; S2: Construct a nonlinear multi-physics field coupling model, including heat conduction field, mechanical field and vibration field. The heat conduction field is modeled based on the thermal conductivity of the material, the thermal effect caused by mechanical work and the external heat source. The mechanical field is modeled based on the nonlinear elastic-plastic relationship and thermal expansion effect. The vibration field is modeled based on the temperature-related stiffness matrix and external dynamic load. S3: Construct a micro-scale model to describe the internal grain sliding behavior, phase change characteristics of the material and its impact on macroscopic multi-physics fields; S4: Through the homogenization method, the micro-scale model is mapped to the macroscopic field, and a multi-field coupling model from micro to macro is established; S5: Dynamically solve the coupled model and optimize the solution process by adjusting the time step and boundary conditions in real time; S6: Real-time monitoring of dynamic responses of multi-physics fields based on stability analysis.

2. A mechanical structure industrial simulation method according to claim 1, characterized in that: In step S2, the heat conduction field is modeled by combining material density, specific heat capacity, thermal conductivity with mechanical work and external heat source, and boundary conditions include the relationship between normal heat flux density and ambient temperature and convection heat transfer coefficient.

3. A mechanical structure industrial simulation method according to claim 1, characterized in that: In step S2, the mechanical field is constructed based on a nonlinear elastic-plastic relationship, which includes the initial yield strength, strain hardening coefficient, strain rate sensitivity and thermal softening effect of the material, and also includes the thermal expansion stress caused by the thermal expansion coefficient and elastic modulus.

4. A mechanical structure industrial simulation method according to claim 1, characterized in that: In step S2, the mechanical field is constructed based on a nonlinear elastic-plastic relationship, which includes the initial yield strength, strain hardening coefficient, strain rate sensitivity and thermal softening effect of the material, and also includes the thermal expansion stress caused by the thermal expansion coefficient and elastic modulus.

5. A mechanical structure industrial simulation method according to claim 1, characterized in that: In step S3, the micro-scale model is modeled based on material grain size, slip resistance and thermal strengthening effect, and is further optimized by describing the change process of phase transformation volume fraction.

6. A mechanical structure industrial simulation method according to claim 1, characterized in that: In step S4, the mapping between the micro-scale model and the macroscopic multi-physical field is achieved through a homogenization method, and the homogenization method is used to convert the microscopic stress distribution into a macroscopic stress component to ensure the continuity of the microscopic effect in the macroscopic field.

7. A mechanical structure industrial simulation method according to claim 1, characterized in that: In step S5, the dynamic solution is combined with the deep learning model, and the neural network is used to accelerate the mapping process from the micro-scale model to the macro-scale model, and realize real-time response under complex working conditions.

8. A mechanical structure industrial simulation system, characterized in that: A mechanical structure industrial simulation method according to any one of claims 1 to 7, the system comprising: An input module for receiving a geometric model, material properties and boundary conditions of a mechanical structure; A modeling module, used to construct a nonlinear multi-physics field coupling model, wherein the model includes a heat conduction field, a mechanical field, and a vibration field; Microscopic modeling module, used to describe the material grain slip behavior, phase change characteristics, and establish microscopic scale models; A mapping module for mapping micro-scale models to macroscopic multiphysics through homogenization methods; Solving module, used for dynamic solving of nonlinear multi-physics coupling models; Optimization module for adjusting simulation parameters based on real-time feedback, including dynamic optimization of time steps, boundary conditions, and stiffness matrices.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method for industrial simulation of a mechanical structure as claimed in any one of claims 1 to 7 is implemented.

10. A readable storage medium, characterized in that: The readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for industrial simulation of a mechanical structure according to any one of claims 1 to 7 is implemented.

Images