Method, device and equipment for generating finite element model of large body casting

Abstract

The application discloses a method, device and equipment for generating a finite element model of a large-scale body casting, comprising the following steps: obtaining a body integrated die casting and a corresponding three-dimensional geometric model; establishing a die casting mold model and a casting system model, and discretizing the models into volume grids; performing mold flow analysis on the die casting mold model and the casting system model to obtain process performance field data of the die casting; performing grid discretization processing and sampling on the three-dimensional geometric model to obtain a tensile sampling point set, and performing tensile testing on the die casting to obtain mechanical performance values; calculating mechanical performance variances of each sub-region of each preset partition strategy according to a preset partition measurement and quantitative mapping relationship prediction model, determining an optimal partition strategy, dividing the initial finite element model according to the optimal partition strategy, and generating a finite element model. The method realizes differentiated assignment of material properties in space, and improves the accuracy of index prediction.

Description

Technical Field

[0001] This invention relates to the field of vehicle engineering technology, and in particular to a method, apparatus and equipment for generating finite element models of large vehicle body castings. Background Technology

[0002] During the high-pressure, high-speed filling and rapid cooling process of large, monolithic die-cast parts, the cooling rates, pressure processes, and melt flow directions vary significantly across different regions, resulting in spatial differences in their microstructure and macroscopic mechanical properties. Traditional finite element modeling methods typically treat the entire die-cast part as a homogeneous material, assigning it a set of average material properties. This deviates significantly from physical reality, leading to substantial deviations in the prediction of stress concentration, deformation modes, and energy absorption characteristics during collision simulations, thus affecting the accuracy and optimization effectiveness of safety designs. Summary of the Invention

[0003] In view of the above, the present invention aims to provide a method, apparatus and equipment for generating finite element models of large automotive body castings, so as to solve the aforementioned technical problems.

[0004] The technical solution adopted in this invention is as follows:

[0005] This invention provides a method for generating finite element models for large automotive body castings, including:

[0006] Obtain the physical integrated die-cast body part and its three-dimensional geometric model;

[0007] Based on the actual die-cast part and the three-dimensional geometric model, a die-casting mold model and a casting system model are established, and both the die-casting mold model and the casting system model are discretized into volume grids.

[0008] By pre-setting die-casting material parameters and pre-setting die-casting mold flow process performance, mold flow analysis is performed on the die-casting mold model and the casting system model to obtain process performance field data of the die-casting parts;

[0009] The three-dimensional geometric model is discretized into a mesh and sampled to obtain a set of stretched sampling points;

[0010] Taking each sampling point of the tensile sampling point set as the center, a quasi-static tensile test is performed on the actual die-cast body part to obtain the mechanical property values ​​of the sampling points;

[0011] Based on the process performance field data and the mechanical performance values, a quantitative mapping relationship prediction model from the process parameter field to the mechanical performance field is established.

[0012] The three-dimensional geometric model is meshed using finite element methods to obtain an initial finite element model.

[0013] Based on the preset partitioning strategy, the mechanical performance variance of each sub-region of each partitioning strategy is calculated through the quantitative mapping relationship prediction model to determine the optimal partitioning strategy.

[0014] The initial finite element model is divided into regions according to the optimal partitioning strategy to generate a finite element model.

[0015] Optionally, mold flow analysis is performed on the die-casting mold model and the gating system model to obtain process parameter field data of the die-casting part, including:

[0016] Based on the fluid volume method, mold flow analysis was performed on the die casting mold model and the casting system model to obtain the process parameter field data of the die casting.

[0017] Optionally, by pre-setting die-casting material parameters and pre-setting die-casting mold flow analysis process performance, mold flow analysis is performed on the die-casting mold model and the casting system model to obtain process performance field data of the die-casting part, including:

[0018] By pre-setting die casting material parameters and pre-setting die casting mold flow analysis, and by performing filling-solidification coupled mold flow analysis on the die casting mold model and casting system model based on the fluid volume method, the process parameter field data of the die casting is obtained.

[0019] The process parameter field data includes pressure distribution, temperature distribution, gas phase distribution, shrinkage cavity distribution, and cooling rate.

[0020] Optionally, establishing a quantitative mapping relationship prediction model from the process parameter field to the mechanical property field also includes: mapping the process parameter field data to the finite element structure mesh, and assigning a corresponding process parameter label to each mesh element.

[0021] Optionally, the three-dimensional geometric model is discretized into a mesh and sampled to obtain a set of stretched sampling points, including:

[0022] The three-dimensional geometric model is discretized into a mesh to obtain a set of surface mesh points;

[0023] Monte Carlo sampling is performed on the surface grid point set to obtain the stretched sampling point set.

[0024] Optionally, based on a preset partitioning strategy, the mechanical performance variance of each sub-region of each partitioning strategy is calculated using the initial finite element model to determine the optimal partitioning strategy, including:

[0025] The initial finite element model is divided into regions based on at least two preset partitioning strategies to obtain multiple sub-regions corresponding to each partitioning strategy.

[0026] The mechanical performance variance of each sub-region under each partitioning strategy is calculated using the quantitative mapping relationship prediction model to obtain the variance distribution set corresponding to each partitioning strategy.

[0027] Multiple variance distribution sets are weighted and sorted, and the partitioning strategy with the smallest weighted result is determined as the optimal partitioning strategy; the preset partitioning strategies include partitioning strategies based on the filling process, partitioning strategies based on the solidification temperature field, and partitioning strategies based on the forming pressure field.

[0028] Optionally, the initial finite element model is divided into regions according to the optimal partitioning strategy to obtain a finite element model, including:

[0029] The initial finite element model is divided into regions according to the optimal partitioning strategy, and the average mechanical properties and average stress-strain curves of each sub-region are predicted.

[0030] The average stress-strain curve is converted into a true stress-effective plastic strain curve and then discretized.

[0031] Each sub-region is assigned independent material properties containing the discretized curves to generate a finite element model with spatially differentiated elastoplastic constitutive parameters suitable for explicit dynamic collision simulation.

[0032] The present invention also provides an apparatus for generating finite element models of large automotive body castings, comprising:

[0033] The acquisition module is used to acquire the physical integrated die-cast body part and the three-dimensional geometric model of the physical integrated die-cast part;

[0034] The processing module is used to establish a die-casting mold model and a casting system model based on the physical die-cast part and the three-dimensional geometric model, and to discretize the die-casting mold model and the casting system model into volumetric meshes; to perform mold flow analysis on the die-casting mold model and the casting system model by preset die-casting material parameters and preset die-casting mold flow process performance to obtain process performance field data of the die-cast part; to perform mesh discretization processing and sampling on the three-dimensional geometric model to obtain a set of tensile sampling points; to perform quasi-static tensile tests on the physical die-cast part of the vehicle body with each sampling point of the tensile sampling point set as the center to obtain the mechanical performance values ​​of the sampling points; to establish a quantitative mapping relationship prediction model from the process parameter field to the mechanical performance field based on the process performance field data and the mechanical performance values; to perform finite element meshing on the three-dimensional geometric model to obtain an initial finite element model; to calculate the mechanical performance variance of each sub-region of each partitioning strategy according to a preset partitioning strategy through the quantitative mapping relationship prediction model to determine the optimal partitioning strategy; and to divide the initial finite element model into regions according to the optimal partitioning strategy to generate a finite element model.

[0035] The present invention also provides a computing device, comprising: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above.

[0036] The present invention also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above.

[0037] The above-described solution of the present invention has at least the following beneficial effects:

[0038] The above-described solution of the present invention involves obtaining a physical integrated die-cast body part and its three-dimensional geometric model; establishing a die-casting mold model and a casting system model based on the physical die-cast part and the three-dimensional geometric model, and discretizing both the die-casting mold model and the casting system model into volumetric meshes; performing mold flow analysis on the die-casting mold model and the casting system model by pre-setting die-casting material parameters and pre-setting die-casting mold flow process performance to obtain process performance field data of the die-cast part; performing mesh discretization processing and sampling on the three-dimensional geometric model to obtain a set of tensile sampling points; and using the tensile sampling points... Centered on each sampling point in the sample set, a quasi-static tensile test is performed on the actual die-cast body part to obtain the mechanical property values ​​of the sampling points. Based on the process performance field data and the mechanical property values, a quantitative mapping relationship prediction model from the process parameter field to the mechanical property field is established. The three-dimensional geometric model is meshed using finite element methods to obtain an initial finite element model. According to a preset partitioning strategy, the mechanical property variance of each sub-region of each partitioning strategy is calculated using the quantitative mapping relationship prediction model to determine the optimal partitioning strategy. The initial finite element model is then partitioned into regions according to the optimal partitioning strategy to generate a new finite element model. This achieves spatially differentiated assignment of material properties, enabling the finite element model to more realistically simulate the mechanical behavior of the die-cast part during collision, significantly improving the accuracy of predictions for key indicators such as failure prediction, deformation mode, and energy absorption characteristics. Attached Figure Description

[0039] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described below with reference to the accompanying drawings, wherein:

[0040] Figure 1 A flowchart illustrating a method for generating a finite element model of a large vehicle body casting, provided in an embodiment of the present invention.

[0041] Figure 2 A schematic diagram of a module for generating finite element models of large vehicle body castings provided in an embodiment of the present invention. Detailed Implementation

[0042] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0043] This invention proposes an embodiment of a method for generating finite element models of large automotive body castings, specifically, as follows: Figure 1 As shown, it includes:

[0044] Step 11: Obtain the physical integrated die-cast body part and its three-dimensional geometric model;

[0045] Step 12: Based on the actual integral die-cast part and the three-dimensional geometric model, establish the die-casting mold model and the casting system model, and discretize the die-casting mold model and the casting system model into volume meshes;

[0046] Step 13: By presetting die casting material parameters and presetting die casting mold flow process performance, perform mold flow analysis on the die casting mold model and the casting system model to obtain process performance field data of the die casting;

[0047] Step 14: Discretize the three-dimensional geometric model into a mesh and sample it to obtain a set of stretched sampling points;

[0048] Step 15: Using each sampling point of the tensile sampling point set as the center, perform a quasi-static tensile test on the actual integrated die-cast body part to obtain the mechanical property values ​​of the sampling points;

[0049] Step 16: Based on the process performance field data and the mechanical performance values, establish a quantitative mapping relationship prediction model from the process parameter field to the mechanical performance field;

[0050] Step 17: Perform finite element mesh generation on the three-dimensional geometric model to obtain the initial finite element model;

[0051] Step 18: Based on the preset partitioning strategy, calculate the variance of mechanical performance of each sub-region of each partitioning strategy using the quantitative mapping relationship prediction model, and determine the optimal partitioning strategy.

[0052] Step 19: Divide the initial finite element model into regions according to the optimal partitioning strategy to generate a finite element model.

[0053] In this embodiment, in step 11, the physical object of the integral die casting and the corresponding three-dimensional geometric model are obtained. The three-dimensional geometric model is a CAD format file with no broken surfaces and no gaps. Here, no broken surfaces means that each surface is a complete polygon without missing edges or broken surfaces. No gaps means that the surfaces fit together completely without gaps or overlaps.

[0054] In step 12, a die-casting mold model is established, including a fixed mold, a moving mold, a slider, inserts, a cooling system, venting channels, overflow channels, etc. The die-casting mold model needs to accurately reflect the position, diameter, and connection relationship of the cooling pipes; a casting system model is established, including a pressure chamber / punch, a sprue sleeve, a constant flow channel / divided flow channel, an inner gate, etc.; the above die-casting mold model and casting system model are discretized into a volume grid, wherein the volume grid is a tetrahedral or hexahedral grid, and local mesh refinement is performed in thin-walled regions.

[0055] In step 13, the parameters of the die-casting material, the die-casting mold material, and the thermal interaction coefficient between the casting and the mold are first set. Here, the die-casting material is specifically set to aluminum alloy, and its properties include thermophysical properties (density, specific heat capacity, thermal conductivity), rheological properties (viscosity, surface tension), phase transformation properties (solidspot temperature, liquidus temperature, latent heat of solidification, solid fraction-temperature curve), and mechanical properties (elastic modulus, Poisson's ratio, coefficient of thermal expansion, yield strength), etc.; the die-casting mold material is specifically set to mold steel, and its properties include density, specific heat capacity, and thermal conductivity.

[0056] Next, mold flow analysis is performed on the die-casting mold model and the gating system model. Here, the die-casting mold flow process performance is set, including slow injection speed, switching position, fast injection speed curve, pressurization pressure value, holding time, initial mold temperature, cooling parameters, spraying parameters, etc.; filling analysis, solidification analysis, and thermal stress coupling analysis based on fluid volume method are selected, and the calculations are performed in the mold flow analysis software to obtain process parameter field data such as pressure distribution cloud map, temperature distribution cloud map, gas phase distribution cloud map, shrinkage cavity distribution cloud map, and cooling rate of the die casting.

[0057] In steps 14 and 15, the surface of the three-dimensional geometric model of the integral die-cast part is discretized into a mesh. The mesh shape is quadrilateral, and the mesh size is larger than the minimum size of the stretched sample. The vertices of the mesh are numbered to obtain a point set. For point sets Monte Carlo sampling was performed to obtain a set of stretched sampling points. .

[0058] stretching the sampling point set Centered on the casting, plate-shaped specimens conforming to the tensile test specimen standard were directly cut from the casting body using wire cutting; a coordinate system was established with the center of the casting gate as the origin to obtain the set of tensile sampling points. The three-dimensional coordinates of all points are obtained; quasi-static tensile tests are performed on plate-shaped specimens to obtain mechanical property values ​​such as yield strength, tensile strength, elongation after fracture, and stress-strain curves.

[0059] Furthermore, a structured database is established, with each record containing a unique identifier for the tensile test sample. For each sample, the record information includes location information: the coordinates of the sampling center point; mechanical property values: yield strength, tensile strength, elongation after fracture, and stress-strain curve; and mold flow analysis prediction values: pressure value, temperature value, gas phase distribution, shrinkage cavity distribution, and cooling rate.

[0060] In step 16, based on the process parameter field data obtained from the mold flow analysis and the mechanical property data obtained from the tensile test of the actual object, two types of quantitative mapping prediction models are constructed: one is a mechanical property scalar prediction model, which is used to predict the three core mechanical indicators of yield strength, tensile strength and elongation after fracture at any position in space, serving the variance calculation of the subsequent optimal partitioning strategy; the other is a stress-strain full curve prediction model, which is used to predict the complete elastoplastic stress-strain constitutive relationship of each region, serving the differentiated material property assignment of the final finite element model.

[0061] (1) For the scalar prediction model of mechanical properties, this embodiment preferably uses a backpropagation (BP) neural network to build a quantitative mapping relationship prediction model from the process parameter field to the mechanical property field. The model construction and training process is as follows:

[0062] Based on the aforementioned structured sample database, the process characteristics and mechanical property labels of all tensile specimens are extracted to construct the model training dataset:

[0063] Input feature vector: Each sample corresponds to a set of 8-dimensional input features:

[0064] ;

[0065] in The three-dimensional spatial coordinates of the sampling center point, This is the molding pressure value. This is the solidification temperature value. This refers to the gas phase volume fraction. This represents the volume fraction of the shrinkage cavity. This refers to the cooling rate.

[0066] Output label vector: Each specimen corresponds to a set of 3D mechanical property scalar labels:

[0067] ;

[0068] in For yield strength, For tensile strength, Elongation after fracture.

[0069] To eliminate the dimensional differences between different physical quantities and improve the model convergence speed, max-min normalization is performed on the input features and output labels respectively, mapping all parameters to the [0,1] interval. The normalization formula is:

[0070] ;

[0071] in , These represent the maximum and minimum values ​​of the corresponding features in all samples.

[0072] After normalization, the sample set is randomly divided into training, validation and test sets in a ratio of 7:2:1, which are used for model parameter updates, generalization performance monitoring and final accuracy verification, respectively.

[0073] The backpropagation (BP) neural network adopts a three-layer fully connected BP neural network structure, including an input layer, a single hidden layer, and an output layer. Each layer sequentially completes feature transfer and nonlinear mapping. The specific operation logic is as follows:

[0074] Input layer: It has 8 neurons, each corresponding to an 8-dimensional input feature. It is only responsible for passing the normalized feature data to the hidden layer and has no additional computation.

[0075] Hidden layer: Consists of 16 neurons, employing the Sigmoid activation function to achieve a nonlinear mapping from process parameters to mechanical features. The hidden layer... The formula for calculating the output of a neuron is:

[0076] ;

[0077] in, For the first The values ​​of each input feature; For the input layer The first neuron is connected to the hidden layer. The connection weights of each neuron; For the hidden layer Bias terms for each neuron; This is the Sigmoid activation function.

[0078] Output layer: Three neurons are used, each corresponding to one of the three mechanical properties. A linear activation function is used to output the prediction results. (Output layer number 1) The formula for calculating the original predicted value of a neuron is:

[0079] ;

[0080] in, For the hidden layer The nth neuron to the output layer The connection weights of each neuron; For the output layer Bias terms for each neuron.

[0081] After the original predicted values ​​are inversely normalized to their actual dimensions, they become the predicted values ​​of yield strength, tensile strength and elongation after fracture, which are the final output of the model.

[0082] The model is iteratively trained using gradient descent combined with backpropagation algorithm. By continuously adjusting network weights and biases, the deviation between predicted and true values ​​is minimized. The specific training steps are as follows:

[0083] Parameter initialization: A normal distribution random initialization method is adopted, assigning small values ​​in the range [-1,1] to all connection weights and bias terms to avoid the activation function entering the saturation region and gradient vanishing due to excessively large initial weights.

[0084] Forward propagation calculation: Training set samples are input into the network in batches, and forward propagation is performed layer by layer according to the above calculation formulas for hidden layers and output layers to obtain the predicted mechanical performance value under the current weight parameters. .

[0085] Loss function calculation: The mean squared error (MSE) is used as the loss function to quantify the difference between the predicted value and the true label. Overall error between:

[0086] ;

[0087] in This represents the number of training samples in the current batch.

[0088] Iterative convergence judgment: Repeat the forward propagation calculation, error backpropagation and parameter update process; after each round of full sample training, use the validation set to calculate the model prediction error. When the validation set error no longer decreases for 20 consecutive rounds, or when the number of training rounds reaches the preset maximum number of iterations (1000 times in this embodiment), training is stopped, and the set of weight parameters with the lowest validation error is saved as the final quantitative mapping relationship prediction model.

[0089] Model accuracy verification: The accuracy of the trained model is verified using an independent test set to ensure that the relative error of the predicted yield strength and tensile strength is less than 3% and the relative error of the predicted elongation after fracture is less than 5%, thus meeting the engineering accuracy requirements of collision simulation.

[0090] (2) For the full stress-strain curve prediction model, extract the stress-strain curve corresponding to each sample from the structured database established in the above steps. The stress-strain curve is a set of two-dimensional coordinate data. Flatten it into a one-dimensional vector and form a matrix. ,in, For the sample size, For the stress-strain curve data dimensions corresponding to the sample, This represents the number of data points on the stress-strain curve.

[0091] According to the following formula, for the matrix Perform eigenvalue decomposition to obtain the principal component score matrix. and principal component matrix :

[0092] ;

[0093] Selecting the principal component matrix The front of the middle Principal component pairs Make an estimate so that Here, T stands for matrix transpose.

[0094] Furthermore, a multiple linear regression model was selected, using the location coordinates of the sampling center point and the pressure value. Temperature value Gas phase volume fraction Volume fraction of shrinkage cavity Cooling rate Using the least squares method as the independent variable, the coefficients in the regression model are estimated and fitted to establish a principal component score matrix regression prediction model for prediction. As shown in the following formula, then combine the formula Obtain a regression prediction model for the stress-strain curve.

[0095] ;

[0096] In the formula, Given a score matrix of k principal components, The coefficient matrix of the fitted regression prediction model is shown below. This is the offset matrix. The feature matrix specifically consists of the three-dimensional position coordinates of the sampling center point and the pressure value. Temperature value Gas phase volume fraction Volume fraction of shrinkage cavity Cooling rate .

[0097] The stress-strain curve regression prediction model can predict a flattened one-dimensional vector based on a set of characteristic parameters, and then recover the two-dimensional coordinate data of the stress-strain curve through matrix transformation, thereby obtaining the predicted stress-strain curve.

[0098] Furthermore, the process parameter field data is mapped to the finite element structure mesh:

[0099] Specifically, firstly, extract the center coordinates of each volumetric grid cell from the model flow analysis results file. and their corresponding pressure values Temperature value Gas phase volume fraction Volume fraction of shrinkage cavity Cooling rate Secondly, the coordinates of the integration points of each shell element in the finite element model of the structure are traversed. Using the nearest neighbor interpolation algorithm, a search is performed within a preset radius. Calculate the equivalent process performance value at the integration point of the shell element based on the mold flow analysis data points within the specified range. The specific calculation formula is as follows:

[0100] ;

[0101] in The technological performance values ​​of the structural mesh points. For the first [unit] within the search radius The process performance values ​​of individual meshes, For spatial distance, Distance weighting factor (in this embodiment, it is taken as...) ).

[0102] Through the above mapping, each mesh element of the initial finite element model obtains corresponding attribute labels for pressure, temperature, gas volume fraction, cooling rate, and shrinkage volume fraction. Then, the corresponding mechanical property values ​​are predicted by the mechanical property scalar prediction model, providing a data basis for subsequent regional variance calculation.

[0103] In step 17, the three-dimensional geometric model of the integral die-cast part is meshed using finite element methods to obtain an initial finite element model with multiple sub-regions. The file format of this initial finite element model is CAE model.

[0104] Next, the initial finite element model is divided into regions, and the specific division strategy is as follows:

[0105] First partition strategy: Specifically, the first partitioning strategy divides the space according to the filling process. Indicates the area near the gate. This indicates the wall thickness zone below the main channel. Indicates the distal thin-walled region. Indicates the area where reinforcing ribs intersect. Indicates the central area of ​​the planar wall. Indicates the overflow channel area;

[0106] Second partitioning strategy: Specifically, the second zoning strategy divides the area according to the temperature distribution after molding. Indicates the area with the highest temperature. This indicates the region with the second highest temperature, and so on. Indicates the area with the lowest temperature;

[0107] Third partition strategy: Specifically, the third partitioning strategy is based on the pressure distribution after molding. This indicates the area of ​​highest pressure. This indicates the area with the second highest pressure, and so on. This indicates the area with the lowest pressure.

[0108] In step 18, the mesh in the initial finite element model is divided into regions according to the first partitioning strategy to obtain the finite element mesh elements of the sub-regions corresponding to the first partitioning strategy. ;

[0109] Extracting finite element mesh elements from sub-regions The coordinates of the center node are used to predict the finite element mesh elements in the sub-region using a quantitative mapping relationship prediction model. Yield strength, tensile strength, and elongation after fracture;

[0110] Calculate all finite element mesh elements in the sub-region sequentially. The variances of yield strength, tensile strength, and elongation after fracture are calculated, and these variances are statistically represented as a set of variance distributions. .

[0111] Repeat the above steps to calculate the variance distribution sets of yield strength, tensile strength, and elongation after fracture for all finite element mesh elements in the sub-regions of the second and third partitioning strategies. , ;

[0112] right , , The partitioning strategy with the smallest variance is the optimal partitioning strategy.

[0113] In step 19, the initial finite element model is divided into regions according to the optimal partitioning strategy to obtain the optimal region mesh elements after optimal partitioning:

[0114] ;

[0115] Calculate all grid cells in the optimal region using a quantitative mapping relationship prediction model. The average yield strength, tensile strength, and elongation after fracture. .

[0116] Applying a stress-strain curve regression prediction model, the optimal region grid element is predicted and calculated. The stress-strain curves were obtained, and the average stress-strain curve values ​​were calculated. .

[0117] Create a separate material card for each sub-region:

[0118] ;

[0119] The average values ​​of yield strength, tensile strength, and elongation after fracture Mean stress-strain curve values Material properties are assigned In the process, high-precision collision simulation verification analysis was conducted to obtain finite element models with differentiated material properties.

[0120] Specifically, the average stress-strain curve value To flatten a one-dimensional vector, matrix operations are used to restore it to a set of engineering stress-strain coordinate points. To ensure that the curve can be correctly read and calculated by finite element solvers (such as LS-DYNA and Abaqus), the following post-processing is performed:

[0121] The engineering stress-strain curve is converted into a true stress-effective plastic strain curve using the following formula: .

[0122] in, Represents the actual stress. and The original data pairs output by the stress-strain curve regression prediction model. This is the elastic modulus constant in the preset die-casting material parameters.

[0123] The transformed curves are resampled at equal intervals to generate a fixed number of points (e.g., 8 or 16 data pairs) that meet the constitutive card requirements of a specific material, and then written into the material card of the corresponding sub-region. middle.

[0124] The method for generating finite element models for large automotive body castings in this embodiment establishes a mapping relationship between process performance and material performance, and performs refined partitioning modeling of the die castings. This enables the material properties to be assigned differential values ​​in space, allowing the finite element model to more realistically simulate the mechanical behavior of the die castings during the collision process. This significantly improves the accuracy of predicting key indicators such as failure prediction, deformation mode, and energy absorption characteristics.

[0125] The refined zoning modeling allows for the prediction of the performance distribution of each region of the die casting based on the preset process performance in the early stages of die casting design, enabling the proactive identification of potential weak areas and the achievement of precise process rectification.

[0126] The refined modeling capability provided by this invention is a key technology connecting physical manufacturing and digital space, and can provide a digital foundation for achieving full life-cycle quality control, performance traceability and intelligent upgrading of large integrated die-cast parts.

[0127] like Figure 2 As shown, a finite element model generation device 20 for large vehicle body castings includes:

[0128] The acquisition module 21 is used to acquire the physical integrated die-cast body part and the three-dimensional geometric model of the physical integrated die-cast part;

[0129] The processing module 22 is used to establish a die-casting mold model and a casting system model based on the actual integrated die-casting part and the three-dimensional geometric model, and to discretize the die-casting mold model and the casting system model into volumetric meshes; to perform mold flow analysis on the die-casting mold model and the casting system model by preset die-casting material parameters and preset die-casting mold flow process performance to obtain process performance field data of the die-casting part; to perform mesh discretization processing and sampling on the three-dimensional geometric model to obtain a set of tensile sampling points; to perform quasi-static tensile testing on the actual integrated die-casting part of the vehicle body with each sampling point of the tensile sampling point set as the center to obtain the mechanical performance values ​​of the sampling points; to establish a quantitative mapping relationship prediction model from the process parameter field to the mechanical performance field based on the process performance field data and the mechanical performance values; to perform finite element meshing on the three-dimensional geometric model to obtain an initial finite element model; to calculate the mechanical performance variance of each sub-region of each partitioning strategy according to the preset partitioning strategy through the quantitative mapping relationship prediction model to determine the optimal partitioning strategy; and to divide the initial finite element model into regions according to the optimal partitioning strategy to generate a finite element model.

[0130] Optionally, mold flow analysis is performed on the die-casting mold model and the gating system model to obtain process parameter field data of the die-casting part, including:

[0131] Based on the fluid volume method, mold flow analysis was performed on the die casting mold model and the casting system model to obtain the process parameter field data of the die casting.

[0132] Optionally, by pre-setting die-casting material parameters and pre-setting die-casting mold flow analysis process performance, mold flow analysis is performed on the die-casting mold model and the casting system model to obtain process performance field data of the die-casting part, including:

[0133] By pre-setting die casting material parameters and pre-setting die casting mold flow analysis, and by performing filling-solidification coupled mold flow analysis on the die casting mold model and casting system model based on the fluid volume method, the process parameter field data of the die casting is obtained.

[0134] The process parameter field data includes pressure distribution, temperature distribution, gas phase distribution, shrinkage cavity distribution, and cooling rate.

[0135] Optionally, establishing a quantitative mapping relationship prediction model from the process parameter field to the mechanical property field also includes: mapping the process parameter field data to the finite element structure mesh, and assigning a corresponding process parameter label to each mesh element.

[0136] Optionally, the three-dimensional geometric model is discretized into a mesh and sampled to obtain a set of stretched sampling points, including:

[0137] The three-dimensional geometric model is discretized into a mesh to obtain a set of surface mesh points;

[0138] Monte Carlo sampling is performed on the surface grid point set to obtain the stretched sampling point set.

[0139] Optionally, based on a preset partitioning strategy, the mechanical performance variance of each sub-region of each partitioning strategy is calculated using the initial finite element model to determine the optimal partitioning strategy, including:

[0140] The initial finite element model is divided into regions based on at least two preset partitioning strategies to obtain multiple sub-regions corresponding to each partitioning strategy.

[0141] The mechanical performance variance of each sub-region under each partitioning strategy is calculated using the quantitative mapping relationship prediction model to obtain the variance distribution set corresponding to each partitioning strategy.

[0142] Multiple variance distribution sets are weighted and sorted, and the partitioning strategy with the smallest weighted result is determined as the optimal partitioning strategy; the preset partitioning strategies include partitioning strategies based on the filling process, partitioning strategies based on the solidification temperature field, and partitioning strategies based on the forming pressure field.

[0143] Optionally, the initial finite element model is divided into regions according to the optimal partitioning strategy to obtain a finite element model, including:

[0144] The initial finite element model is divided into regions according to the optimal partitioning strategy, and the average mechanical properties and average stress-strain curves of each sub-region are predicted.

[0145] The average stress-strain curve is converted into a true stress-effective plastic strain curve and then discretized.

[0146] Each sub-region is assigned independent material properties containing the discretized curves to generate a finite element model with spatially differentiated elastoplastic constitutive parameters suitable for explicit dynamic collision simulation.

[0147] It should be noted that this device is the same as the method described above. All implementations in the above method embodiments are applicable to the embodiments of this device and can achieve the same technical effect.

[0148] An embodiment of the present invention also provides a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described in the above embodiments. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0149] In this embodiment of the invention, a computer-readable storage medium is also provided, storing instructions that, when executed on a computer, cause the computer to perform the method described in the above embodiments. All implementations of the methods described in the above embodiments are applicable to this embodiment and can achieve the same technical effect.

[0150] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0151] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0152] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0153] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0154] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0155] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0156] Furthermore, it should be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of the present invention. Moreover, the steps performing the above series of processes can naturally be executed in the order described, but are not necessarily required to be executed in chronological order; some steps can be executed in parallel or independently of each other. Those skilled in the art will understand that all or any step or component of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or a combination thereof. This is something that those skilled in the art can achieve by using their basic programming skills after reading the description of the present invention.

[0157] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing device. The computing device can be a known general-purpose device. Therefore, the object of the present invention can also be achieved simply by providing a program product containing program code implementing the method or apparatus. That is, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any known storage medium or any storage medium developed in the future. It should also be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent to the present invention. Furthermore, the steps performing the above series of processes can naturally be performed in the order described, but are not necessarily required to be performed in chronological order. Some steps can be performed in parallel or independently of each other.

[0158] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for generating a finite element model for large automotive body castings, characterized in that, include: Obtain the physical integrated die-cast body part and its three-dimensional geometric model; Based on the actual die-cast part and the three-dimensional geometric model, a die-casting mold model and a casting system model are established, and both the die-casting mold model and the casting system model are discretized into volume grids. By pre-setting die-casting material parameters and pre-setting die-casting mold flow process performance, mold flow analysis is performed on the die-casting mold model and the casting system model to obtain process performance field data of the die-casting parts; The three-dimensional geometric model is discretized into a mesh and sampled to obtain a set of stretched sampling points; Taking each sampling point of the tensile sampling point set as the center, a quasi-static tensile test is performed on the actual die-cast body part to obtain the mechanical property values ​​of the sampling points; Based on the process performance field data and the mechanical performance values, a quantitative mapping relationship prediction model from the process parameter field to the mechanical performance field is established. The quantitative mapping relationship prediction model includes a mechanical performance scalar prediction model and a stress-strain full curve prediction model. The establishment of a quantitative mapping relationship prediction model from process parameter field to mechanical property field further includes: mapping process parameter field data to finite element structure mesh, and assigning a corresponding process parameter label to each mesh element; The three-dimensional geometric model is meshed using finite element methods to obtain an initial finite element model. Based on the preset partitioning strategy, the mechanical performance variance of each sub-region of each partitioning strategy is calculated through the quantitative mapping relationship prediction model to determine the optimal partitioning strategy. The initial finite element model is divided into regions according to the optimal partitioning strategy to generate a finite element model; Specifically, by pre-setting die-casting material parameters and pre-setting die-casting mold flow process performance, mold flow analysis is performed on the die-casting mold model and the casting system model to obtain process performance field data of the die-casting part, including: By pre-setting die casting material parameters and pre-setting die casting mold flow analysis, and by performing filling-solidification coupled mold flow analysis on the die casting mold model and casting system model based on the fluid volume method, the process parameter field data of the die casting is obtained. The process parameter field data includes pressure distribution, temperature distribution, gas phase distribution, shrinkage cavity distribution, and cooling rate; Specifically, based on a preset partitioning strategy, the mechanical performance variance of each sub-region of each partitioning strategy is calculated using the quantitative mapping relationship prediction model to determine the optimal partitioning strategy, including: The initial finite element model is divided into regions based on at least two preset partitioning strategies to obtain multiple sub-regions corresponding to each partitioning strategy. The mechanical performance variance of each sub-region under each partitioning strategy is calculated using the quantitative mapping relationship prediction model to obtain the variance distribution set corresponding to each partitioning strategy. Multiple variance distribution sets are weighted and sorted, and the partitioning strategy with the smallest weighted result is determined as the optimal partitioning strategy; the preset partitioning strategies include partitioning strategies based on the filling process, partitioning strategies based on the solidification temperature field, and partitioning strategies based on the forming pressure field.

2. The method for generating a finite element model for large vehicle body castings according to claim 1, characterized in that, Mold flow analysis was performed on the die-casting mold model and the gating system model to obtain process parameter field data of the die-casting part, including: Based on the fluid volume method, mold flow analysis was performed on the die casting mold model and the casting system model to obtain the process parameter field data of the die casting.

3. The method for generating a finite element model for large vehicle body castings according to claim 1, characterized in that, The three-dimensional geometric model is discretized into a mesh and sampled to obtain a set of stretched sampling points, including: The three-dimensional geometric model is discretized into a mesh to obtain a set of surface mesh points; Monte Carlo sampling is performed on the surface grid point set to obtain the stretched sampling point set.

4. The method for generating a finite element model for large vehicle body castings according to claim 1, characterized in that, The initial finite element model is divided into regions according to the optimal partitioning strategy to obtain the finite element model, including: The initial finite element model is divided into regions according to the optimal partitioning strategy, and the average mechanical properties and average stress-strain curves of each sub-region are predicted. The average stress-strain curve is converted into a true stress-effective plastic strain curve and then discretized. Each subregion is assigned independent material properties containing discretized curves to generate a finite element model with spatially differentiated elastoplastic constitutive parameters for explicit dynamic collision simulation.

5. A device for generating finite element models for large automotive body castings, characterized in that, include: The acquisition module is used to acquire the physical integrated die-cast body part and the three-dimensional geometric model of the physical integrated die-cast part; The processing module is used to establish a die-casting mold model and a casting system model based on the physical die-casting part and the three-dimensional geometric model, and to discretize the die-casting mold model and the casting system model into volume grids. By pre-setting die-casting material parameters and pre-setting die-casting mold flow process performance, mold flow analysis is performed on the die-casting mold model and the casting system model to obtain process performance field data of the die-casting parts; The three-dimensional geometric model is discretized into a mesh and sampled to obtain a set of tensile sampling points; with each sampling point in the set of tensile sampling points as the center, a quasi-static tensile test is performed on the actual die-cast body to obtain the mechanical property values ​​of the sampling points; Based on the process performance field data and the mechanical performance values, a quantitative mapping relationship prediction model from the process parameter field to the mechanical performance field is established. This quantitative mapping relationship prediction model includes a mechanical performance scalar prediction model and a stress-strain full-curve prediction model. The establishment of this model further includes: mapping the process parameter field data to a finite element structure mesh, assigning a corresponding process parameter label to each mesh element; performing finite element mesh generation on the three-dimensional geometric model to obtain an initial finite element model; calculating the mechanical performance variance of each sub-region of each partitioning strategy using the quantitative mapping relationship prediction model according to a preset partitioning strategy, and determining the optimal partitioning strategy; and dividing the initial finite element model into regions according to the optimal partitioning strategy to generate a finite element model. Specifically, by pre-setting die-casting material parameters and pre-setting die-casting mold flow process performance, mold flow analysis is performed on the die-casting mold model and the casting system model to obtain process performance field data of the die-casting part. This includes: analyzing process performance by pre-setting die-casting material parameters and pre-setting die-casting mold flow, and performing filling-solidification coupled mold flow analysis on the die-casting mold model and the casting system model based on the fluid volume method to obtain process parameter field data of the die-casting part; the process parameter field data includes pressure distribution, temperature distribution, gas phase distribution, shrinkage cavity distribution, and cooling rate. The process includes: calculating the mechanical property variance of each sub-region under each zoning strategy based on a preset zoning strategy using the quantitative mapping relationship prediction model, and determining the optimal zoning strategy. This includes: dividing the initial finite element model into regions based on at least two preset zoning strategies to obtain multiple sub-regions corresponding to each zoning strategy; calculating the mechanical property variance of each sub-region under each zoning strategy using the quantitative mapping relationship prediction model to obtain a set of variance distributions corresponding to each zoning strategy; weighting and sorting the multiple variance distribution sets, and determining the zoning strategy with the smallest weighted result as the optimal zoning strategy. The preset zoning strategies include zoning strategies based on the filling process, zoning strategies based on the solidification temperature field, and zoning strategies based on the forming pressure field.

6. A computing device, characterized in that, include: A processor, a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The system stores instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 4.

Images