Metal composite forming die stress optimization system based on finite element analysis

Abstract

The invention relates to the field of computer aided design, in particular to a metal composite forming die stress optimization system based on finite element analysis. A stress change coefficient a caused by the change of the model grid and a stress change coefficient b caused by the change of the die-casting parameter are obtained when the value of the die-casting parameter is changed every two adjacent times and the model grid is adjusted, and the change quantity of the die-casting parameter is in positive correlation with b when the die-casting parameter is changed every time; the grid density adjustment amount is positively correlated with a when the model grid is adjusted each time. And after the values of the die-casting parameters are changed for multiple times, the die-casting parameters when the stress distribution is minimum serve as optimized die-casting parameters. According to the method, the model grid distribution and the values of the die casting parameters are adjusted in real time, so that the finite element analysis process is more efficient.

Description

Technical Field

[0001] The present invention relates to the field of computer-aided design, and particularly to a stress optimization system for a metal composite forming die based on finite element analysis. Background Art

[0002] Finite element analysis is a common method for stress optimization of metal composite forming dies. It creates a three-dimensional model based on computer-aided design software, then simulates and predicts the stress distribution based on this three-dimensional model, and performs stress optimization. The conventional stress optimization method based on finite element analysis is to set different optimization parameters, continuously perform finite element analysis on the three-dimensional model under multiple optimization parameters, and then determine the optimal optimization parameters.

[0003] However, in the process of continuously performing multiple finite element analyses, due to the inappropriate model grid distribution of the three-dimensional model, the efficiency of the finite element analysis stress optimization process is low (that is, the computing power consumption is large, the calculation time is long, and the stress optimization result may not be optimal either); for example, when the model grid is locally too dense, the computing power consumption of the finite element analysis is large and the calculation time is long. If the model grid is locally too sparse, the predicted stress distribution may have obvious errors. More notably, when there are more optimization parameters set, the efficiency brought by the inappropriate model grid distribution described in the above process will be even lower, resulting in the inability to efficiently obtain comprehensive and accurate optimal optimization parameters, and further resulting in the stress optimization result obtained from the entire finite element analysis process not being optimal. Summary of the Invention

[0004] To solve the above problems, the present invention provides a stress optimization system for a metal composite forming die based on finite element analysis.

[0005] The stress optimization system for a metal composite forming die based on finite element analysis of the present invention adopts the following technical solutions: An embodiment of the present invention provides a stress optimization system for a metal composite forming die based on finite element analysis, and the system includes the following modules: A finite element analysis module, configured to perform finite element analysis on the model of the die and the die-casting parameters to obtain a stress distribution, and the position points where the stress in the stress distribution is greater than and not greater than a preset threshold th are respectively defined as target points and non-target points; A model grid adjustment module, configured to change the values of the die-casting parameters and adjust the model grid multiple times; Wherein, the stress distribution and model obtained by changing the die-casting parameters and adjusting the model grid for the i-th time are respectively denoted as 、 ; the model obtained by adjusting the model grid for the (i + 1)-th time is denoted as ; For the model Perform finite element analysis on the die-casting parameters after the (i + 1)-th change to obtain the stress distribution , obtain relative to an area where the target points increase; the stress change coefficient a caused by the change in the model mesh is positively correlated with the correlation between the stress increase amount and the grid density increase amount in the said area; the stress change coefficient b caused by the change in the die-casting parameters is positively correlated with the correlation between the stress increase amount and the grid density decrease amount in the said area; When the difference between a and b is greater than the preset difference, after changing the value of th, recalculate a and b using the model mesh adjustment module; When changing the die-casting parameters for the (i + 2)-th time, make the die-casting parameter change amount positively correlated with b. The (i + 2)-th adjustment of the model mesh includes: increasing the grid density at the target points in , the grid density adjustment amount is positively correlated with a, and reducing the grid density at the non-target points in; The die-casting parameter optimization module, after changing the values of the die-casting parameters multiple times, takes the die-casting parameters when the stress distribution is the smallest as the optimized die-casting parameters.

[0006] Preferably, the step of increasing the grid density at the target points in includes the following specific steps: On the model , obtain the grids where all the target points in are located, denoted as the first grids, denote the number of the first grids as K1, denote the sum of the areas of all the first grids as S1, denote the grid density adjustment amount when increasing the grid density at the target points in as , which means increasing grids per unit area on average. Let , and split each first grid into N1 grids, where N1 is rounded to the nearest integer.

[0007] Preferably, the step of reducing the grid density at the non-target points in includes the following specific steps: On the model , obtain the grids where all the non-target points in are located, denoted as the second grids, denote the number of the second grids as K2, denote the sum of the areas of all the second grids as S2, denote the reduction amount of the grid density when reducing the grid density at the non-target points as , let , and N2 = K2 - , where is rounded to the nearest integer, and reduce the K2 second grids to N2 grids; The value is equal to the grid density adjustment amount.

[0008] Preferably, the stress change coefficient a caused by the change of the model grid is positively correlated with the correlation between the stress increase amount and the grid density increase amount in the region, and the specific steps are as follows: For any position point in the region where the target point increases, obtain the stress of this position point in , obtain the stress of this position point in , and use as the stress increase amount of this position; obtain the grid density of this position point on , obtain the grid density of this position point on , and use as the grid point density increase amount of this position point; For all position points where the grid point density increase amount is greater than the preset percentage, the Pearson correlation coefficient between the stress increase amounts of all position points in the region where the target point increases and the grid point density increase amounts is denoted as the first correlation; the product of the first correlation and the first preset parameter k1 is denoted as a.

[0009] Preferably, the stress change coefficient b caused by the change of the die-casting parameters is positively correlated with the correlation between the stress increase amount and the grid density decrease amount in the region, and the specific steps are as follows: For any position point in the region where the target point increases, obtain the stress of this position point in , obtain the stress of this position point in , and use as the stress increase amount of this position; obtain the grid density of this position point on , obtain the grid density of this position point on , and use as the grid density decrease amount of this position point; For all position points where the grid point density increase amount is greater than the preset percentage, the Pearson correlation coefficient between the stress increase amounts of all position points in the region where the target point increases and the grid density decrease amounts is denoted as the second correlation, and the product of the second correlation and the second preset parameter k2 is denoted as b.

[0010] Preferably, when the difference between a and b is greater than the preset difference, change the value of th, and the specific steps are as follows: When the difference between a and b is greater than the preset difference, if the number of target points is greater than the number of non-target points, then increase the value of th, and the increased th is th + ft, where ft represents the change amount of th; if the number of target points is less than or equal to the number of non-target points, then decrease the value of th, and the decreased th is th - ft.

[0011] Preferably, the specific steps for changing the die-casting parameters for the (i + 2)-th time are as follows: The die-casting parameters obtained after changing the die-casting parameters for the (i + 1)-th time are denoted as , and the die-casting parameters are regarded as the parameter points composed of the pouring temperature, injection speed, and holding pressure time. Taking the die-casting parameters as the center and the change amount of the die-casting parameters as the radius, randomly select a parameter point on the spherical surface as the die-casting parameters after the (i + 2)-th change.

[0012] Preferably, the change amount of the die-casting parameters is b×R, where R is a preset value.

[0013] Preferably, the grid density adjustment amount is: min1 + a×(max1 - min1), where min1 represents the preset minimum value of the grid density adjustment amount, and max1 represents the preset maximum value of the grid density adjustment amount.

[0014] Preferably, the change amount of th is negatively correlated with |a - b|.

[0015] The beneficial effects of the technical solution of the present invention are as follows: On the one hand, the present invention adjusts the distribution of the model grid in real time and dynamically. By increasing the grid density at the target points in the stress distribution (i.e., at the positions with higher stress), the accuracy of the finite element analysis results in the stress abnormal area is improved, so that in subsequent stress optimization (i.e., in the die-casting parameter optimization module), it can be avoided that the stress abnormal area (such as the area with larger stress) cannot be detected in time, resulting in the problem that the die-cast parts still have stress abnormal areas after stress optimization. By reducing the grid density at non-target points (i.e., at the positions with lower stress), the problem of excessive grid in the stress normal area leading to large computing power consumption and long calculation time can be avoided. On the other hand, the value of the die-casting parameters is adjusted in real time and dynamically, which can avoid wasting too much computing power and calculation time in the finite element analysis process corresponding to multiple changes of the die-casting parameters in a local small range when stress abnormal areas are likely to appear; when stress abnormal areas are not likely to appear, the die-casting parameters can be adjusted more finely and continuously to ensure that the die-casting parameters at the optimal stress can be obtained comprehensively as a whole subsequently. At this time, at the cost of sacrificing computing power and calculation time, the problem that the optimal die-casting parameters cannot be obtained due to large changes in the die-casting parameters can be avoided.

[0016] Furthermore, the present invention dynamically adjusts the value of th, so that when the values of die-casting parameters are changed multiple times and the model grid is adjusted, it is ensured that the division of target points and non-target points is more reasonable, so that it can be distinguished or clarified whether the increase in target points is caused by the change of the model grid or the change of die-casting parameters, and it is avoided that the positions or regions that do not belong to abnormal stress in actual production are regarded as the positions or regions of abnormal stress (that is, divided into target points), or the positions or regions that belong to abnormal stress in actual production are not regarded as the positions or regions of abnormal stress (that is, not divided into target points), thereby resulting in relatively inaccurate optimized die-casting parameters and relatively inefficient calculation processes (for example, more wasted computing power and longer calculation time). BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0018] Figure 1 It is a framework structure diagram of a stress optimization system for a metal composite forming die based on finite element analysis provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of a die model of a stress optimization system for a metal composite forming die based on finite element analysis provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following will, in conjunction with the drawings and preferred embodiments, detail the specific implementation manners, structures, features and effects of the stress optimization system for a metal composite forming die based on finite element analysis proposed by the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0021] The following will specifically describe the specific solution of the stress optimization system for a metal composite forming die based on finite element analysis provided by the present invention with reference to the drawings.

[0022] Embodiment 1: Please refer to Figure 1, which shows the framework structure diagram of the stress optimization system for metal composite forming dies based on finite element analysis provided by an embodiment of the present invention. The system includes the following modules: a finite element analysis module, a model mesh adjustment module, and a die-casting parameter optimization module. The system runs on a computer device.

[0023] The finite element analysis module is used to predict the stress distribution of the model through finite element analysis based on the three-dimensional model of the metal composite forming die and the die-casting parameters of the die.

[0024] The model mesh adjustment module is used to change the values of the die-casting parameters and adjust the model mesh multiple times. Each time the die-casting parameters are changed, the model mesh is adjusted once and a finite element analysis is performed again. On the one hand, it avoids obvious errors in the predicted stress caused by unreasonable model mesh distribution (for example, the mesh in some areas of the model is too dense and in some areas is too sparse); on the other hand, it avoids the problem that the model mesh occupies or wastes too much computing power during the process of changing the values of the die-casting parameters multiple times, resulting in too long calculation or the inability to perform more finite element analyses in a short time, thereby ensuring that stress optimization can be carried out from more finite element analysis processes subsequently.

[0025] The die-casting parameter optimization module is used to obtain the optimal die-casting parameters, make them suitable for the current metal composite forming die, and make it have the optimal stress distribution.

[0026] Specifically: The finite element analysis module includes the following steps: Step S101: Use computer-aided design software to create a three-dimensional model of the metal composite forming die.

[0027] In this embodiment, computer-aided design software (such as CAD) is used to design the three-dimensional model of the die (hereinafter referred to as the model). In other embodiments, Maya, 3dMax, etc. can be used to design the three-dimensional model of the die. This embodiment does not make specific limitations. Figure 2 shows a model of a die, which is designed using CAD.

[0028] The model in this embodiment is composed of triangular mesh elements and mesh points. In other embodiments, the model can be divided into hexahedron or tetrahedron elements. In this embodiment, HyperMesh software is used to divide the model mesh. CAD, Maya, HyperMesh, etc. are all existing computer-aided design software. The principles and operation methods of this embodiment are not described.

[0029] It should be noted that the above-mentioned software related to computer-aided design has been installed in the computer device in advance.

[0030] Step S102: Based on the model of the die and the die-casting parameters of the die, predict the stress distribution of the model through finite element analysis.

[0031] Import the model of the die (such as a model in STEP or IGES format) into finite element analysis software (such as ANSYS SpaceClaim computer-aided design software), and at the same time set the material properties and boundary conditions.

[0032] As an example, when die-casting zinc alloy, the material properties and boundary conditions refer to the specific heat capacity, thermal conductivity, solid / liquidus temperature, etc. of zinc alloy and the elastic modulus, Poisson's ratio, etc. of die steel. In this embodiment, the die-casting die uses SKD61 hot-work die steel. These materials have good hot hardness, wear resistance and thermal fatigue resistance, and can meet the requirements of zinc alloy die-casting.

[0033] The die-casting parameters in this embodiment refer to three index data: pouring temperature, injection speed, and holding pressure time. In some other embodiments, when die-casting is carried out in stages, the number of stages, the pouring temperature, injection speed, holding pressure time and other index data in each stage can be used as die-casting parameters together.

[0034] In one embodiment, the initial values of the die-casting parameters are: pouring temperature 400 °C, injection speed 70 m / s, and holding pressure time 1.2 s. In other embodiments, the initial values of the die-casting parameters can be set to other values, which are not specifically limited here. In this embodiment, in order to eliminate the interference of the dimension and order of magnitude of the die-casting parameters, the initial values in the die-casting parameters in this embodiment are set as: the ratio of the pouring temperature 400 °C to the maximum pouring temperature 420 °C, the ratio of the injection speed 70 m / s to the maximum injection speed 80 m / s, and the ratio of the holding pressure time 1.2 s to the maximum holding pressure time 3 s.

[0035] Based on the model of the die, die-casting parameters, and material properties and boundary conditions, simulate the die-casting process of zinc alloy through finite element analysis software, and predict the stress distribution of the die at the mold opening stage. The stress distribution described is the stress magnitude at each position point on the model surface.

[0036] In the stress distribution of this embodiment, the position points where the stress is greater than the preset threshold th are defined as target points, and the position points where the stress is less than or equal to the preset threshold th are non-target points. This process divides the stress distribution into two parts for subsequent analysis and processing. One part has relatively high stress (i.e., target points), and these stresses are more likely to cause springback or deformation of the die-cast parts; the other part has relatively low stress (i.e., non-target points), and these stresses are less likely to cause springback or deformation of the die-cast parts. In this embodiment, th = 40 MPa is used as an example for description. In other embodiments, th can be set to other values, which are not specifically limited in this embodiment.

[0037] As an example, the method for obtaining each position point on the model surface is as follows: The model is in a three-dimensional space of computer-aided three-dimensional design software (such as CAD, Maya, 3DMax). The three-dimensional space is evenly divided into cubic grids with equal volumes. In this embodiment, the volume of each cubic grid is 1 cubic millimeter. In other embodiments, the volume of each cubic grid can be set to other values, which is not specifically limited in this embodiment. The center point of the cubic grid passed by the grid of the model is recorded as a position point of the grid of the model (this position point is a three-dimensional coordinate point). In other embodiments, the center point of the model grid within each cubic grid can also be used as the position point of the grid of the model.

[0038] As another example, the method for obtaining each position point on the model surface is as follows: Since the grid of the model is the surface of the model, using the method of model UV unwrapping, the grid of the three-dimensional model is flattened and rendered into a two-dimensional image. In this embodiment, the size of the flattened and rendered image is 2048×2048. At this time, each pixel point on this image represents each position point on the model surface (this position point is a two-dimensional coordinate point). Model UV unwrapping is a well-known method and will not be specifically described in this embodiment.

[0039] The model grid adjustment module is used to change the values of die-casting parameters and adjust the model grid multiple times, specifically including: Step S201: Denote the initial die-casting parameters as (that is, the initial value of the die-casting parameters described in step S102), denote the initial model as (that is, the model obtained in step S101), and denote the initial stress distribution as (that is, the stress distribution obtained in step S102).

[0040] For the convenience of subsequent narration, denote as the value of the die-casting parameters after the 0th change of the die-casting parameters; denote as the model obtained after the 0th adjustment of the model grid; Denote

[0041] as the stress distribution obtained after the 0th change of the die-casting parameter value and the adjustment of the model grid. 、 、 The process of obtaining

[0042] The process of the 1st change of the die-casting parameter value and the adjustment of the model grid (abbreviated as the 1st process) includes: At Based on the change of the value of , the obtained result is denoted as , representing the value of the die-casting parameters after the first change of the die-casting parameters; Adjust the mesh density on . The specific method is as follows: for the target points and non-target points in (the specific method is shown in step S102), increase the mesh density at the target points and decrease the mesh density at the non-target points on

[0043] Then, perform finite element analysis on the model of the mold and the die-casting parameters of the mold to predict the stress distribution of the mold in the mold-opening stage , representing the stress distribution obtained after the first change of the value of the die-casting parameters and the adjustment of the model mesh.

[0044] The process of the second change of the value of the die-casting parameters and the adjustment of the model mesh (abbreviated as the second process) includes: On the basis of , change the value of , and the obtained result is denoted as , representing the value of the die-casting parameters after the second change of the die-casting parameters; Adjust the mesh density on . The specific method is as follows: for the target points and non-target points in (the specific method is shown in step S102), increase the mesh density at the target points and decrease the mesh density at the non-target points on

[0045] Then, perform finite element analysis on the model of the mold and the die-casting parameters of the mold to predict the stress distribution of the mold in the mold-opening stage , representing the stress distribution obtained after the second change of the value of the die-casting parameters and the adjustment of the model mesh.

[0046] And so on, perform the process of the i-th change of the value of the die-casting parameters and the adjustment of the model mesh (abbreviated as the i-th process). The stress distribution and the model obtained after the i-th change of the die-casting parameters and the adjustment of the model mesh are respectively denoted as , , and the obtained die-casting parameters are denoted as .

[0047] Then, the process of changing the values of die-casting parameters and adjusting the model mesh for the (i + 1)-th time (abbreviated as the (i + 1)-th process) is as follows: On the basis of change the value, and the obtained result is denoted as , representing the value of the die-casting parameters after the (i + 1)-th change of the die-casting parameters; Adjust the mesh density of the model . The specific method is: for the target points and non-target points in (see the specific method in step S102), increase the mesh density at the target points and decrease the mesh density at the non-target points on , and the obtained model is denoted as , representing the model obtained after the (i + 1)-th adjustment of the model mesh.

[0048] Then, perform finite element analysis on the model of the mold and the die-casting parameters of the mold to predict the stress distribution of the mold in the mold opening stage , representing the stress distribution obtained after the (i + 1)-th change of the die-casting parameter values and adjustment of the model mesh.

[0049] Subsequently, continue to do so by analogy, and this embodiment will not elaborate further.

[0050] As an example, the process of changing the values of die-casting parameters and adjusting the model mesh is set to 50 times, that is, stop after the 50th change of the values of die-casting parameters and adjustment of the model mesh. In other embodiments, 50 here can be replaced with other values, and this embodiment does not make specific limitations.

[0051] As another example, limit the process of changing the values of die-casting parameters and adjusting the model mesh multiple times within 8 hours, that is, end after the steps of the model mesh adjustment module are executed for 8 hours.

[0052] The specific problem considered above is: in finite element analysis, the mesh density of the model significantly affects the accuracy of the stress distribution prediction result. Among them, increasing the mesh density can significantly improve the accuracy of the stress distribution, but when the mesh density is too high, it will increase the computational amount (i.e., computing power) during finite element analysis. Especially in this embodiment, the values of die-casting parameters need to be changed multiple times, and a finite element analysis needs to be performed each time the die-casting parameters are changed. If the mesh density is too high, there will be a huge computational amount and it will take a long time.

[0053] In the above process of this step, the mesh density at the target points in the stress distribution is increased on the model, and the mesh density at non-target points is reduced to achieve real-time adjustment of the model density (or real-time redistribution of the model mesh distribution); by increasing the mesh density at the target points in the stress distribution (i.e., at the positions with higher stress), the accuracy of the finite element analysis results in the stress abnormal area is improved, so that in subsequent stress optimization (i.e., in the die-casting parameter optimization module), it can be avoided that the stress abnormal area (such as the area with larger stress) cannot be detected in time, resulting in the problem that the die-cast parts still have stress abnormal areas after stress optimization. By reducing the mesh density at non-target points (i.e., at the positions with lower stress), the problem of excessive mesh in the stress normal area leading to large computing power consumption and long calculation time can be avoided.

[0054] In summary, the above process requires multiple finite element analyses when optimizing stress. In multiple finite element analyses, by reasonably redistributing the density distribution of the model mesh, while ensuring the high efficiency of the finite element analysis (i.e., reducing large computing power consumption and long calculation time), the situation that the stress abnormal area cannot be optimized and removed due to stress prediction errors is avoided.

[0055] Step S202: For the process of changing the values of die-casting parameters and adjusting the model mesh twice in succession, such as the process of changing the values of die-casting parameters and adjusting the model mesh at the i-th and (i + 1)-th times, the following processing is carried out: (1) First of all, it should be noted that when performing finite element analysis in the i-th and (i + 1)-th processes, the die-casting parameters and the model mesh density used are different. Therefore, and the distribution of target points may be different, that is, the target points in change relative to the target points in

[0056] First, obtain the area where the target points increase relative to and denote it as , that is, the target points in area in increase compared with the target points in this area in

[0057] (2) Then estimate the stress change coefficient a caused by the change of the model mesh, and estimate the stress change coefficient b caused by the change of the die-casting parameters.

[0058] It should be noted that in this step, considering the increase in the number of target points (equivalent to the increase in the number of stress anomaly position points), on the one hand, it is caused by the change of the model grid, and on the other hand, it is caused by the die-casting parameters; among them, the larger the stress change coefficient a, the more likely it is caused by the change of the model grid; the larger the stress change coefficient b, the more likely it is caused by the change of the die-casting parameters.

[0059] Furthermore, it should be noted that if within the area where the target points increase (i.e., within the area ), the increase in stress is caused by the increase in grid density, it indicates that the stress change coefficient a is larger; if within the area where the target points increase (i.e., within the area ), the grid density decreases but the stress increases significantly, then it indicates that the stress change coefficient b is larger.

[0060] Therefore, a is positively correlated with the first correlation, and the first correlation refers to the correlation between the increase in stress and the increase in grid density within the area where the target points increase (i.e., within the area ); b is positively correlated with the second correlation, and the second correlation refers to the correlation between the increase in stress and the decrease in grid density within the area where the target points increase (i.e., within the area ).

[0061] Specifically, when the area does not exist, let a = b = 0.25.

[0062] (3) After obtaining the change coefficient a and the change coefficient b, the process of changing the value of the die-casting parameters and adjusting the model grid for the (i + 2)-th time (referred to as the (i + 2)-th process) can be executed, including: On the basis of change the value, and the obtained result is denoted as , representing the value of the die-casting parameters after the (i + 2)-th change of the die-casting parameters. Among them, when changing to , the change amount of is positively correlated with b.

[0063] Adjust the grid density on the model . The specific method is: for the target points and non-target points in , increase the grid density at the target points and decrease the grid density at the non-target points on , and the obtained model is denoted as , representing the model obtained after the (i + 2)-th adjustment of the model grid. Among them, when increasing the grid density at the target points on , the adjustment amount of the grid density is positively correlated with a.

[0064] In this process, for The change in is positively correlated with b, indicating that if the previous changes in die casting parameters (i.e., the i-th and i+1-th processes) are more likely to cause or cause an increase in target points (i.e., when abnormal stress areas are more likely to occur), then a larger change is required in the i+2 process. , to avoid wasting too much computing power and time on the corresponding finite element analysis process when changing the die-casting parameters multiple times in a small local area when stress abnormal areas are prone to appear. If the previous changes in the die-casting parameters (i.e., the i-th and i+1-th processes) are less likely to cause or cause the increase of target points, then a smaller change is required in the i+2 process. , so that when stress abnormal areas are not likely to appear, the die-casting parameters can be adjusted more finely and continuously to ensure that the die-casting parameters with optimal stress can be obtained in an overall and comprehensive manner in the future. At this time, the problem of the optimal die-casting parameters not being able to be obtained due to a significant change in the die-casting parameters is avoided at the expense of computing power and calculation time.

[0065] The adjustment amount of the grid density is positively correlated with a, which means: if the increase of target points is more likely to be caused when the model grid density is adjusted (the i-th and i+1-th adjustment of the model grid density), then the grid density needs to be increased more significantly, that is, when the stress abnormality area is likely to appear, further redistribution is performed to increase the increase of the grid density. The purpose is to improve the accuracy of stress prediction by increasing the grid density of the model, so that the target points and non-target points can be distinguished more accurately in the future, so that the grid density can be allocated or adjusted more reasonably in the future. If the increase of target points is less likely to be caused when the model grid density is changed (equivalent to no obvious stress abnormality area), then the grid density of the model needs to be increased less significantly. The reason is: while avoiding wasting too much computing power and calculation time, the unreasonable grid density will not be caused by the existence of large stress prediction errors when the grid density is redistributed or adjusted in the future. The unreasonable grid density refers to: the low grid density in the area with high stress (or abnormal stress) leads to the inability to further accurately predict stress, and the high grid density in the area with low stress (or normal stress) leads to the waste of computing power and calculation time in finite element analysis.

[0066] At this point, the model grid adjustment module is summarized as follows: H1: The model mesh adjustment module continuously changes the values ​​of die-casting parameters and adjusts the model mesh, and continuously obtains stress distribution. Each process performs a finite element analysis, and each die-casting parameter corresponds to a stress distribution.

[0067] H2: Continuously changing the values of die-casting parameters and adjusting the model grid is a recursive process, that is, the process of changing the values of die-casting parameters and adjusting the model grid next time is based on the die-casting parameters, model, and stress distribution obtained in the previous time (after changing the values of die-casting parameters and adjusting the model grid).

[0068] H3: Each time the values of die-casting parameters are changed and the model grid is adjusted, the change amount of die-casting parameters and the increase amount of the model grid are obtained from the change coefficient a and the change coefficient b, where the change coefficient a and the change coefficient b are obtained from the previous two processes of changing the values of die-casting parameters and adjusting the model grid; For example, the change coefficient a and the change coefficient b used in the (i + 2)-th process are obtained from the i-th and (i + 1)-th processes. That is: The change coefficient a and the change coefficient b used in the 2-nd process are obtained from the 0-th and 1-st processes; the change coefficient a and the change coefficient b used in the 3-rd process are obtained from the 1-st and 2-nd processes, and so on.

[0069] Specifically, when the values of die-casting parameters are changed multiple times, if the values of each index data in the die-casting parameters exceed the upper and lower boundaries of the expected value range, then the values of the index data are set to the values corresponding to the upper and lower boundaries of the value range. For example, the pouring temperature is 390 - 420 °C, the injection speed is 30 - 80 m / s, and the holding pressure time is 0.5 - 3 s.

[0070] The die-casting parameter optimization module, the specific method of this module is: After changing the values of die-casting parameters multiple times, the die-casting parameters when the stress distribution is the smallest are used as the optimized die-casting parameters.

[0071] The optimized die-casting parameters describe that: for the mold provided by the finite element analysis module, when die-casting is carried out using this mold, die-casting is carried out using the optimized die-casting parameters, so that the obtained die-cast part has a smaller stress, and the die-cast part can meet the production requirements. For example, when the die-cast part is integrally formed with other metal parts, the composite formed product will not be unqualified due to the area with excessive stress in the die-cast part.

[0072] It should be noted that the optimized die-casting parameters are dimensionless and without order of magnitude. It is necessary to multiply the index data such as pouring temperature, injection speed, and holding pressure time in the optimized die-casting parameters by the maximum pouring temperature, maximum injection speed, and maximum holding pressure time respectively, so as to obtain the die-casting parameters with dimension and order of magnitude.

[0073] As an example, the smallest stress distribution means: Obtain the mean value of the stresses at all position points in the stress distribution, and regard the stress distribution with the smallest mean value as the smallest stress distribution.

[0074] As another example, the minimum stress distribution means that: Obtain the mean stress of all target points, and regard the stress distribution with the minimum mean value as the minimum stress distribution.

[0075] Thus, the first embodiment ends.

[0076] The present invention provides a comparative embodiment, and the stress optimization method included in this comparative embodiment is: Manually set different die-casting parameters (die-casting parameters refer to pouring temperature, injection speed, and holding pressure time). For example, randomly select 50 die-casting parameters within a pouring temperature of 390 - 420 °C, an injection speed of 30 - 80 m / s, and a holding pressure time of 0.5 - 3 s. Respectively use each die-casting parameter and the model of the same mold for finite element analysis to obtain the stress distribution of each die-casting parameter, and finally use the die-casting parameter when the stress distribution is the smallest as the optimized die-casting parameter.

[0077] This comparative embodiment is an existing or conventional method for stress optimization. The problems of this method are: when the distribution of the manually set die-casting parameters is unreasonable or the number of die-casting parameters is small, it is difficult to obtain optimized die-casting parameters comprehensively and globally. If the number of manually set die-casting parameters is large, it will face huge computational workload and calculation time, especially when each die-casting parameter contains more index dimensions. Generally speaking, it is difficult for this comparative embodiment to efficiently obtain relatively optimal die-casting parameters.

[0078] Compared with the comparative embodiment, on the one hand, by adjusting the distribution of the model grid in real time and dynamically, by increasing the grid density at the target points in the stress distribution (i.e., the positions with higher stress), the accuracy of the finite element analysis results in the stress abnormal area is improved, so that subsequent stress optimization (i.e., in the die-casting parameter optimization module) can avoid the problem that the stress abnormal area (such as the area with larger stress) cannot be detected in time, resulting in the problem that the die-cast parts still have stress abnormal areas after stress optimization. By reducing the grid density at non-target points (i.e., the positions with lower stress), the problem of large computing power consumption and long calculation time caused by too many grids in the normal stress area can be avoided. On the other hand, the value of the die-casting parameter is adjusted in real time and dynamically, which can avoid wasting too much computing power and calculation time in the finite element analysis process corresponding to changing the die-casting parameter multiple times in a local small range when stress abnormal areas are likely to appear; when stress abnormal areas are not likely to appear, the die-casting parameter can be adjusted more finely and continuously to ensure that the die-casting parameter when the stress is optimal can be obtained comprehensively and globally subsequently. At this time, at the cost of sacrificing computing power and calculation time, the problem that the optimal die-casting parameter cannot be obtained due to a large change in the die-casting parameter is avoided.

[0079] Embodiment 2: This embodiment is a further improvement on the model grid adjustment module in Embodiment 1, specifically an improvement on the calculation process of the change coefficient a and the change coefficient b.

[0080] Reviewing Embodiment 1, it can be seen that: if the increase in stress in the area where the target points increase is due to the increase in grid density, it indicates that the stress change coefficient a is larger; if the grid density decreases but the stress increases significantly in the area where the target points increase, it indicates that the stress change coefficient b is larger. Therefore, a is positively correlated with the first correlation, where the first correlation refers to the correlation between the stress increase amount and the grid density increase amount in the area where the target points increase; b is positively correlated with the second correlation, where the second correlation refers to the correlation between the stress increase amount and the grid density decrease amount in the area where the target points increase.

[0081] However, the above process in Embodiment 1 may be inaccurate. The so-called inaccurate situation means that: the increase in stress in the area where the target points increase, and the decrease in grid density but the significant increase in stress in the area where the target points increase. These two points may be caused by both the change of the model grid and the change of the die-casting parameters at the same time. At this time, it is difficult to distinguish or clarify whether the increase in the target points is caused by the change of the model grid or the change of the die-casting parameters. At this time, the process of changing the values of the die-casting parameters and adjusting the model grid in Embodiment 1 is unreliable (that is, the optimized die-casting parameters obtained in Embodiment 1 are not optimal and the calculation process is not efficient).

[0082] This embodiment takes into account that when the difference between a and b is large, it is very easy to distinguish whether the increase in the target points is caused by the change of the model grid or the change of the die-casting parameters. For example, when a is larger than b, it means that the increase in the target points is more obviously caused by the change of the model grid. Even if a part of it is caused by the change of the die-casting parameters, its influence is small and can be ignored. That is, at this time, implementing according to the method in Embodiment 1 is reliable (that is, the optimized die-casting parameters obtained in Embodiment 1 are relatively optimal and the calculation process is relatively efficient).

[0083] When the difference between a and b is less than the preset difference, it indicates that the difference between a and b is not significant, suggesting that it is difficult to distinguish or clarify whether the increase in the target points is caused by a change in the model grid or a change in the die-casting parameters. The reason may be that the preset threshold th set in the first embodiment is inaccurate. For example, when the preset threshold th is set too large or too small, it will cause a position or area that does not belong to abnormal stress in actual production to be regarded as a position or area of abnormal stress (that is, classified as a target point), or it will cause a position or area that belongs to abnormal stress in actual production not to be regarded as a position or area of abnormal stress (that is, not classified as a target point). As a result, it is difficult to distinguish or clarify whether the increase in the target points between a and b obtained in the first embodiment is caused by a change in the model grid or a change in the die-casting parameters, and further leads to the relatively inaccurate optimized die-casting parameters and relatively inefficient calculation process (for example, more wasted computing power and longer calculation time) obtained in the first embodiment.

[0084] In this embodiment, when it is detected that the difference between a and b is less than the preset difference, the value of th is changed, and then a and b are recalculated using the model grid adjustment module again. For the specific steps, refer to step (2) of step S202 in the first embodiment. In this embodiment, the preset difference is taken as 0.18 for description, and in other embodiments, it can be set to other values, which are not specifically limited in this embodiment.

[0085] It should be noted that during the process of recalculating a and b using the model grid adjustment module, the changed value of th needs to be used to divide the target points and non-target points.

[0086] For example, when changing the value of the die-casting parameter and adjusting the model grid for the (i + 2)-th time, if the difference between a and b is greater than the preset difference, then change the value of th, and then re-obtain the area where the target points increase ; where the method for obtaining the area where the target points increase is the same as that of step S202 of the model grid adjustment module in the first embodiment, and specifically includes: Re-obtain the target points and non-target points in based on the changed th (see step S102 of the first embodiment for details), re-obtain the target points and non-target points in based on the changed th, and then obtain relative to the area where the target points increase, which is denoted as .

[0087] After recalculating a and b, continue to change the value of the die-casting parameter and adjust the model grid according to step (3) of step S202 in the first embodiment; for example, after recalculating a and b, change to , and for The change amount is positively correlated with the recalculated b; when adjusting the model to in the process, when increasing the grid density at the target point on , the adjustment amount of the grid density is positively correlated with the recalculated a.

[0088] As an example, when it is detected that the difference between a and b is less than the preset difference threshold, change the value of th, and the methods included are: When the difference between a and b is greater than the preset difference, if the number of target points is greater than the number of non-target points, then increase the value of th, and the increased th is th + ft, where ft represents the change amount of th. If the number of target points is less than or equal to the number of non-target points, then decrease the value of th, and the decreased th is th - ft.

[0089] As an alternative example, the method for obtaining the change amount ft of th is: set ft to 0.1×th.

[0090] As a preferred example, the method for obtaining the change amount ft of th is: set ft to 0.5 - |a - b|×th, where |a - b| represents the difference between a and b, and the smaller this difference is, the larger ft is, indicating that th needs to be changed more significantly.

[0091] Specifically, when ft is less than or equal to 0, make ft equal to 0, and when ft is greater than 0.5×th, make ft = 0.5×th.

[0092] Specifically, if there are no target points or non-target points in this embodiment, no longer calculate a and b but directly change the value of th, and at this time, the change amount ft of th is set to 0.1×th.

[0093] In this embodiment, by dynamically adjusting the value of th, when changing the values of die-casting parameters and adjusting the model grid multiple times, it is ensured that the division of target points and non-target points is more reasonable, so that it can be distinguished or clarified whether the increase in target points is caused by the change of the model grid or the change of die-casting parameters, and it is avoided that positions or regions that do not belong to abnormal stress in actual production are regarded as positions or regions of abnormal stress (that is, divided into target points), or positions or regions that belong to abnormal stress in actual production are not regarded as positions or regions of abnormal stress (that is, not divided into target points). Furthermore, it avoids the problems that the optimized die-casting parameters obtained in the first embodiment are relatively inaccurate and the calculation process is relatively inefficient (for example, more computing power is wasted and the calculation time is longer).

[0094] Embodiment 3: In this embodiment, on the basis of Embodiment 2, the following methods are further included: As an example, after recalculating a and b, the further steps include: when the difference between a and b is still less than the preset difference threshold, then recalculate a and b again, and so on, until the difference between a and b is greater than or equal to the preset difference, and then continue to change the values of the die-casting parameters and adjust the model grid according to step (3) of step S202 in Embodiment 1.

[0095] Further, when the difference between a and b is still less than the preset difference after several times of repeated calculation of a and b (for example, after 10 times), at this time, obtain the values of a and b when the difference between a and b is the largest among these several times, and for the value of th changed when taking these values of a and b, recalculate a and b again using the model grid adjustment module.

[0096] Embodiment 4: As described in Embodiment 1, when performing the process of changing the values of the die-casting parameters and adjusting the model grid for the (i + 2)-th time, the steps included are: (1) On the basis of change the value of , and the obtained result is denoted as . Among them, when changing to , the change amount of is positively correlated with b.

[0097] (2) For the target points and non-target points in , increase the grid density at the target points and decrease the grid density at the non-target points on , and the obtained model is denoted as . Among them, when increasing the grid density at the target points on , the adjustment amount of the grid density is positively correlated with a.

[0098] In this embodiment, as an example, the method of increasing the grid density at the target points on includes: On the model , obtain the grids where all target points are located, denoted as the first grids, that is, each first grid contains at least one target point. Denote the number of the first grids as K1, and denote the sum of the areas of all the first grids as S1. Use the following method to increase the grid density at the target points: Define a parameter , which represents the adjustment amount of the grid density in the process of the (i + 2)-th adjustment of the model grid, that is, the average number of grids to be increased per unit area is . Then represents: it is necessary to expand the K1 first grids by grids. Then each first grid needs to be split into a grid (N1 needs to be rounded up).

[0099] As an example, the method of splitting each first grid into N1 grids is as follows: Randomly generate N1 position points within each first grid, and use the Thiessen polygon algorithm to generate N1 grids based on the N1 position points.

[0100] As another example, the method of splitting each first grid into N1 grids is as follows: Randomly generate N1×100 position points within each first grid, use the K-Means clustering algorithm to cluster the N1×100 position points into N1 categories, obtain the center points of all position points in each category to get N1 center points, and use the Thiessen polygon algorithm to generate N1 grids based on the N1 center points. In other embodiments, 100 described here can be replaced with other values, and this embodiment does not specifically limit it.

[0101] In this embodiment, as an example, in reduce the grid density at non-target points, and the methods included are as follows: In the model obtain all the grids where non-target points are located, denoted as the second grids, that is, each second grid contains at least one non-target point. Denote the number of second grids as K2, and denote the sum of the areas of all second grids as S2. Use the following method to reduce the grid density at non-target points: Define a parameter , representing the reduction amount of grid density in the process of adjusting the model grid for the (i + 2)-th time, that is, the average reduction of grids per unit area. Then represents that: it is necessary to reduce K2 second grids by grids ( needs to be rounded up), then the reduced number of grids is N2 = K2 - , that is, it is necessary to reduce K2 second grids to N2 grids.

[0102] In this embodiment , where represents the scaling coefficient. This embodiment describes it with pp = 1 as an example, that is, the reduction amount of grids per unit area is equal to the increase amount of grids per unit area. In other embodiments, pp can also be set to other values, such as 1.2 or 0.8, and this embodiment does not specifically limit it.

[0103] As an example, reducing K2 second grids to N2 grids includes: Obtain all the center points of the K2 second grids to get K2 center points. Use the K-Means clustering algorithm to cluster the K2 center points into N2 categories. Each category contains several grids. Merge all the grids in each category into one grid. For example, delete the common edges and fixed points of all the grids in each category, and then merge all the grids in each category into one grid.

[0104] Specifically, when N2 > K2 × 0.5, it means that too many grids are reduced. At this time, let N2 = K2 × 0.5 (rounded to the nearest integer). In other embodiments, 0.5 of this number can be set to other values, such as 1.0. This embodiment does not make specific limitations.

[0105] In the above process of this embodiment, by increasing the grid density at the target point and reducing the grid density at non-target points, the obtained model is denoted as .

[0106] Furthermore, in this embodiment, as an example, when increasing the grid density at the target point on , the adjustment amount of the grid density is positively correlated with a, including: where the adjustment amount of the grid density is also the parameter defined above in this embodiment ; min1 and max1 are two different preset parameters greater than 0, and max1 is greater than min1. In this embodiment, a is greater than or equal to 1 and less than or equal to 0. At this time represents the minimum value, and in this embodiment, taking as an example for description, max1 represents the maximum value, and in this embodiment, max1 = 5 is taken as an example for description.

[0107] In other embodiments, min1 and max1 can be set to other values. This embodiment does not make specific limitations.

[0108] In this embodiment, as an example, when changing to , the change amount of is positively correlated with b, including: Considering that contains multiple indicators, such as pouring temperature, injection speed, and holding pressure time. Regard all the indicators as a vector (for example, a three-dimensional vector composed of pouring temperature, injection speed, and holding pressure time). At this time is represented as a vector in the space composed of all the indicators, that is, a parameter point in the space.

[0109] Centered on the parameter point represented by , randomly select a parameter point on the spherical surface with a radius of b×R. This parameter point serves as 's value. Here, R is a preset value. b×R represents the change amount of . In this embodiment, R = 0.1 is taken as an example for description. In other embodiments, R can be set to other values, as long as it is greater than 0 and less than 1.

[0110] In the above steps of this embodiment, taking the process of changing the values of die-casting parameters and adjusting the model grid for the (i + 2)-th time as an example, a specific method for adjusting the model grid density on the model is given, that is, the specific methods involved in increasing the grid density at the target point, the specific methods involved in reducing the grid density at non-target points, the calculation method of the change amount of , and the calculation method of the increase amount of grid density

[0111] It should be emphasized again that, as described in H3 in the model grid adjustment module of Embodiment 1, the change coefficient a and the change coefficient b used in the (i + 2)-th process are obtained from the i-th and (i + 1)-th times, that is the change amount of and the increase amount of grid density need to rely on the i-th and (i + 1)-th processes. Specifically, the change coefficient a and the change coefficient b used in the first process cannot be obtained, or rather, the change coefficient a and the change coefficient b used in the first process do not need to be obtained. Therefore, in the first process the change amount of

[0112] and the increase amount of grid density cannot be obtained using the above methods in this embodiment. Then, in this embodiment, in the process of changing the values of die-casting parameters and adjusting the model grid for the first time, the method for obtaining the change amount of and the increase amount of grid density is as follows: Set the change amount of to 0.5×R, that is, centered on the parameter point represented by

[0113] randomly select a parameter point on the spherical surface with a radius of 0.5×R. This parameter point serves as 's value, that is, the changed value of .

[0114] Specifically, when the target point is exactly at a grid point, the method for determining the grid where the target point is located is as follows: For all the grids sharing this grid point, obtain the center points of these grids, and take the grid with the closest Euclidean distance between the center point and the target point as the grid where the target point is located.

[0115] Embodiment Five: As described in Embodiment One, during the process of changing the values of the die-casting parameters and adjusting the model grid for the i-th and (i + 1)-th times, and are obtained, and is obtained with respect to There is an area where the target points increase, denoted as ; The stress change coefficient a caused by the change of the model grid is estimated, and the stress change coefficient b caused by the change of the die-casting parameters is estimated. a is positively correlated with the first correlation, and the first correlation refers to the correlation between the stress increase amount and the grid density increase amount within the area where the target points increase (that is, within the area ); b is positively correlated with the second correlation, and the second correlation refers to the correlation between the stress increase amount and the grid density decrease amount within the area where the target points increase (that is, within the area ).

[0116] In this embodiment, as an example, a is positively correlated with the first correlation, and the first correlation refers to the correlation between the stress increase amount and the grid density increase amount within the area where the target points increase (that is, within the area ), and the method included is as follows: For any position point within the area where the target points increase, obtain the stress of this position point in , obtain the stress of this position point in

[0117] and take as the stress increase amount of this position. For any position point within the area where the target points increase, obtain the grid density of this position point on and take

[0118] as the grid point density increase amount of this position point.

[0119] As another example, a is positively correlated with the first correlation, where the first correlation refers to the correlation between the increase in stress and the increase in mesh density within the region where the target point increases (i.e., within the region ), and the method included is as follows: Only consider the position points where the increase in mesh point density is greater than 1%. The Pearson correlation coefficient between the increase in stress at these position points within the region where the target point increases and the increase in mesh point density at all positions is denoted as the first correlation.

[0120] Specifically, if there are no position points where the increase in mesh point density is greater than 1%, then replace 1% with 0.9%. If there are no position points where the increase in mesh point density is greater than 0.9%, then replace 0.9% with 0.8%, and so on until there are some.

[0121] In this embodiment, as an example, b is positively correlated with the second correlation, where the second correlation refers to the correlation between the increase in stress and the decrease in mesh density within the region where the target point increases (i.e., within the region ), and the method included is as follows: For any position point within the region where the target point increases, obtain the stress at this position point in , obtain the stress at this position point in , and use as the increase in stress at this position point.

[0122] For any position point within the region where the target point increases, obtain the mesh density at this position point on , obtain the mesh density at this position point on , and use as the decrease in mesh density at this position point.

[0123] The Pearson correlation coefficient between the increase in stress at all positions within the region where the target point increases and the decrease in mesh density at all positions is denoted as the second correlation. The product of the second correlation and the second preset parameter k2 is denoted as b (note that when b is less than 0.01, set b to 0.01). In this embodiment, k2 = 1 is taken as an example for description, and in other embodiments, k2 can be set to other values, which is not specifically limited in this embodiment.

[0124] As an example, b is positively correlated with the second correlation, where the second correlation refers to the correlation between the increase in stress and the decrease in mesh density within the region where the target point increases (i.e., within the region ), and the method included is as follows: Only consider the position points where the reduction in grid density is greater than 1%. The Pearson correlation coefficient between the increase in stress at these positions within the area where the target points increase and the reduction in grid density at all positions is denoted as the second correlation coefficient.

[0125] Specifically, if there are no position points where the reduction in grid density is greater than 1%, then replace 1% with 0.9%. If there are no position points where the reduction in grid density is greater than 0.9%, then replace 0.9% with 0.8%, and so on until there are some.

[0126] As an embodiment, obtain the grid density at this position point , and the method includes: Obtain the grid where this position point is located. All grids adjacent to this grid are denoted as the first-order neighborhood grids of this grid (excluding this grid). All grids adjacent to the first-order neighborhood grids are denoted as the second-order neighborhood grids of this grid (excluding the first-order neighborhood grids). All grids adjacent to the second-order neighborhood grids are denoted as the third-order neighborhood grids of this grid (excluding the second-order neighborhood grids), and so on. In this embodiment, obtain the grid set composed of this grid, all its first-order neighborhood grids, and all its second-order neighborhood grids. The ratio of the number of grids in the grid set to the total area of the grids is denoted as the grid density , representing the average number of grids per unit area.

[0127] Specifically, when the average grid density of all position points within the grids where all target points are located is greater than the first preset density value, the grid density at all target points will no longer increase to avoid excessive grid density. In this embodiment, the first preset density value is equal to 1 grid per square millimeter. In other embodiments, the first preset density value can be set to other values, and this embodiment does not make specific limitations. Similarly, when the average grid density of all position points within the grids where all non-target points are located is less than the second preset density value, the grid density at all non-target points will no longer decrease to avoid too small grid density. In this embodiment, the second preset density value is equal to 0.00001 grid per square millimeter. In other embodiments, the first preset density value can be set to other values, and this embodiment does not make specific limitations.

[0128] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention shall be included in the protection scope of the present invention. It should be noted that.

Claims

1. A metal composite forming die stress optimization system based on finite element analysis, characterized in that: The system includes the following modules: A finite element analysis module is used to perform finite element analysis on the mold model and die-casting parameters to obtain stress distribution. Position points where the stress is greater than or less than a preset threshold value th in the stress distribution are defined as target points and non-target points, respectively; Model grid adjustment module, used to change the values ​​of die-casting parameters and adjust the model grid multiple times; Among them, the stress distribution and model obtained by changing the die casting parameters and adjusting the model mesh for the i-th time are respectively recorded as , ; The model obtained by adjusting the model grid for the i+1th time is recorded as ; Model And the stress distribution is obtained by finite element analysis of the die casting parameters after the i+1th change , get Relative to There is an area with increased target points; the stress variation coefficient a caused by the change in the model mesh is positively correlated with the correlation between the stress increase and the mesh density increase in the area; the stress variation coefficient b caused by the change in the die-casting parameters is positively correlated with the correlation between the stress increase and the mesh density decrease in the area; When the difference between a and b is greater than the preset difference, the value of th is changed and a and b are recalculated using the model grid adjustment module; When the die casting parameters are changed for the i+2th time, the change amount of the die casting parameters is positively correlated with b. The i+2th adjustment of the model grid includes: Increase The grid density at the target point in the grid density adjustment is positively correlated with a, and the reduction The grid density at non-target points in ; The die-casting parameter optimization module changes the values ​​of the die-casting parameters for multiple times and takes the die-casting parameters with the smallest stress distribution as the optimized die-casting parameters.

2. The metal composite forming die stress optimization system based on finite element analysis according to claim 1, characterized in that: Said in Increase The specific steps of calculating the mesh density at the target point in are as follows: In the model On, get The grid where all target points are located is recorded as the first grid, the number of first grids is recorded as K1, and the sum of the areas of all first grids is recorded as S1. The mesh density adjustment amount at the target point in is recorded as , which means the average increase per unit area grid, , split each first grid into N1 grids, where N1 is rounded to the nearest integer.

3. The metal composite forming die stress optimization system based on finite element analysis according to claim 1, characterized in that: The reduction The specific steps of calculating the mesh density at non-target points in are as follows: In the model On, get The grids where all non-target points are located are recorded as the second grids, the number of second grids is recorded as K2, the sum of the areas of all second grids is recorded as S2, and the reduction in grid density when reducing the grid density at the non-target points is recorded as ,make , and N2=K2- ,in Round off to the nearest integer, reducing K2 second grids to N2 grids; The value of is equal to the mesh density adjustment amount.

4. The metal composite forming die stress optimization system based on finite element analysis according to claim 1, characterized in that: The stress variation coefficient a caused by the model grid change is positively correlated with the correlation between the stress increase in the region and the grid density increase, and the specific steps included are as follows: For any position point in the area where the target point is added, obtain The stress at this point in , get The stress at this point in ,Will As the stress increase at that location; obtain The grid density at this point on , get The grid density at this point on ,Will As the increase in the grid point density at that location; For all position points where the increase in grid point density is greater than a preset percentage, the Pearson correlation coefficient between the stress increase and the grid point density increase of all position points in the area where the target point is increased is recorded as the first correlation; the product of the first correlation and the first preset parameter k1 is recorded as a.

5. The metal composite forming die stress optimization system based on finite element analysis according to claim 1, characterized in that: The stress variation coefficient b caused by the die casting parameter change is positively correlated with the correlation between the stress increase and the mesh density reduction in the region, and the specific steps included are as follows: For any position point in the area where the target point is added, obtain The stress at this point in , get The stress at this point in ,Will As the stress increase at that location; obtain The grid density at this point on , get The grid density at this point on ,Will as the amount by which the mesh density is reduced at that location; For all positions where the increase in grid point density is greater than a preset percentage, the Pearson correlation coefficient between the stress increase and the grid density decrease for all positions in the area where the target point is increased is recorded as the second correlation, and the product of the second correlation and the second preset parameter k2 is recorded as b.

6. The metal composite forming die stress optimization system based on finite element analysis according to claim 1, characterized in that: When the difference between a and b is greater than a preset difference, the value of th is changed, and the specific steps include the following: When the difference between a and b is greater than the preset difference, if the number of target points is greater than the number of non-target points, then the value of th is increased, and the increased th is th+ft, where ft represents the change in th; if the number of target points is less than or equal to the number of non-target points, then the value of th is reduced, and the reduced th is th-ft.

7. The metal composite forming die stress optimization system based on finite element analysis according to claim 1, characterized in that: The i+2th time of changing the die casting parameters includes the following specific steps: The die casting parameters obtained after the i+1th change of the die casting parameters are recorded as The die casting parameters are regarded as the parameter points composed of pouring temperature, injection speed and holding time. Centered on A parameter point is randomly selected on the spherical surface with a radius of the die-casting parameter change as the die-casting parameter after the i+2th change.

8. The metal composite forming die stress optimization system based on finite element analysis according to claim 1 or 7, characterized in that: The die-casting parameter change amount is b×R, where R is a preset value.

9. The metal composite forming die stress optimization system based on finite element analysis according to any one of claims 1, 2 or 3, characterized in that: The grid density adjustment amount is: min1+a×(max1-min1), min1 represents the preset minimum value of the grid density adjustment amount, and max1 represents the preset maximum value of the grid density adjustment amount.

10. The metal composite forming die stress optimization system based on finite element analysis according to claim 6, characterized in that: The change in th is negatively correlated with |ab|.

Images