A simulation method for part milling process considering multi-step stress release process

Through the construction and redistribution of part models, mapping and recombining stress, and cyclic simulation calculation, the multi-step stress release process simulation problem in part milling processing is solved, and precise control and cost reduction of part deformation is achieved.

CN118690615BActive Publication Date: 2024-11-15CHINA AIRPLANT STRENGTH RES INST
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411162053.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-23
Publication Date
2024-11-15
Estimated Expiration
2044-08-23

AI Technical Summary

Technical Problem

The prior art lacks simulation methods for multi-step stress release processes during part milling processing, which makes it difficult to determine the number of stress releases and the impact of material removal on the final deformation during the processing, and consumes a lot of manpower and material resources.

Method used

A simulation method for the part milling process considering the multi-step stress release process is adopted. By constructing part models, obtaining initial residual stress, geometric redistribution and mesh redistribution, mapping stress, and recombining the unit stiffness matrix, the cyclic simulation calculation is carried out until The part model reaches a predetermined thickness to achieve deformation simulation of the multi-step stress relief process.

Benefits of technology

Effectively control the processing deformation of parts, minimize the manpower and material costs of process optimization, improve simulation analysis efficiency, and reduce costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118690615B_ABST
    Figure CN118690615B_ABST
Patent Text Reader

Abstract

The present application belongs to the field of mechanical processing technology, and particularly relates to a method for simulating a part milling process taking into account a multi-step stress release process. The simulation method comprises: constructing a part model, and performing preliminary geometric division and mesh division on the part model; after obtaining the initial residual stress, performing simulation calculation to realize stress release; after each execution of stress release, geometrically re-dividing and meshing the part model, and the geometric re-dividing divides the part model into a milled part and a part part; removing the mesh units of the milled part by reorganizing the unit stiffness matrix to obtain a new part model; then simulating and calculating the part model after milling; and simulating the part milling process through multiple cycles. The present application realizes deformation simulation of parts processed by a multi-step stress release process.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application belongs to the field of mechanical processing technology, and in particular relates to a method for simulating a part milling process taking into account a multi-step stress release process. Background Art

[0002] When processing large metal parts such as thin-walled frames or long beams, the deformation of the parts is significantly affected by the processing technology, especially when the material removal rate is high, and the parts usually have large deformation after processing.

[0003] In engineering, multi-step stress release processes such as stress-free faceting or unequal height process bosses can be used to control part deformation. Specifically, when fine processing is performed after rough processing, the parts are not directly pressed, but unequal height gaskets or process bosses are used according to the deformation of the parts to make the parts in a state without external stress during fine processing, thereby reducing the processing deformation of the parts. However, in actual use, it is difficult to determine the number of times stress release is performed during the processing process and the impact of the amount of material removed during each processing on the final deformation. Generally, it is necessary to explore the process through experience or a large number of experiments, which consumes a lot of manpower and material resources. In addition, the use of simulation methods can effectively improve analysis efficiency and reduce costs. The existing technology uses life and death unit technology to simulate milling deformation, but this method usually does not take into account the multi-step stress release of the process. At the same time, it is limited by the one-time division of the model grid, and cannot consider the process of removing the material along the processing surface after stress release deformation.

[0004] The above reasons lead to the fact that there is currently no simulation method for the stress-free surface finishing process during part milling.

[0005] Therefore, it is desired to have a technical solution to overcome or at least alleviate at least one of the above-mentioned defects of the prior art. Summary of the invention

[0006] The purpose of the present application is to provide a method for simulating a part milling process taking into account a multi-step stress release process, so as to solve at least one problem existing in the prior art.

[0007] The technical solution of this application is:

[0008] A method for simulating a part milling process considering a multi-step stress release process, comprising:

[0009] Step S1, constructing a part model, and performing geometric division and mesh division on the part model;

[0010] Step S2, obtaining initial residual stress, applying the initial residual stress to the part model, and performing simulation calculation to obtain residual stress and deformation;

[0011] Step S3, dividing the part model into a milled part and a part part by geometric redivision;

[0012] Step S4, re-meshing the milled part and the component part respectively;

[0013] Step S5, mapping the residual stress obtained by simulation calculation to the mesh units of the milled part and the component part after mesh re-division, to obtain the residual stress of the milled part and the component part after mesh re-division;

[0014] Step S6, removing the mesh elements of the milled part in the overall stiffness matrix by reorganizing the element stiffness matrix to obtain a part model of the retained part;

[0015] Step S7, applying the residual stress of the component obtained in step S5 to the component model obtained in step S6, performing simulation calculation to obtain the residual stress and deformation, and returning to step S3;

[0016] Step S8, looping steps S3 to S7 until the part model reaches a predetermined thickness.

[0017] In at least one embodiment of the present application, in step S1, constructing a part model includes:

[0018] Obtain material information and clamping boundary conditions of parts;

[0019] Build part models based on part material information and clamping boundary conditions.

[0020] In at least one embodiment of the present application, in step S2, obtaining the initial residual stress and applying the initial residual stress to the part model includes:

[0021] Obtain the initial residual stress through testing;

[0022] The initial residual stress is applied layer by layer to the part model, where the residual stress of each layer is obtained by linear interpolation method:

[0023] ;

[0024] ;

[0025] ;

[0026] ;

[0027] in, E e is the residual stress of the corresponding layer, H is the height of the interpolation layer, H k is the height of the k-th floor, H k-1is the height of the k-1th floor, A is the cross-sectional area, and M is the moment;

[0028] The interpolation layer height is:

[0029] H=Cd

[0030] Where d is the milling depth and C is a positive integer.

[0031] In at least one embodiment of the present application, in step S3, the subdivision plane when geometrically subdividing the part model is determined by:

[0032] h-hs>ha

[0033] Among them, h is the thickness of the part, ha is the maximum deformation of the part, and hs is the distance from the sectioning surface to the bottom of the part.

[0034] In at least one embodiment of the present application, in step S5, the residual stress obtained by simulation calculation is mapped to the mesh units of the milled part and the part after mesh re-division, so as to obtain the residual stress of the milled part and the part after mesh re-division, including:

[0035] Step S51: Select a reference node on the mesh before mesh re-division p 0 ( x i ,y i ,z i );

[0036] Select the target node on the remeshed mesh p ( x,y,z );

[0037] Step S52, determining the influence domain of each target node in the mesh before mesh re-division, wherein a spherical area with the target node as the sphere center and R as the radius is taken as the influence domain;

[0038] Search in the influence domain corresponding to each target node:

[0039] If there is a reference node located at the center of the influence domain, the stress of the reference node located at the center of the influence domain is used as the stress of the target node;

[0040] If there is no reference node located at the center of the influence domain, the reference node inside the influence domain is used as the neighboring node;

[0041] Step S53: Calculate the spatial correlation between the target node and the neighboring nodes according to the spatial correlation function:

[0042] ;

[0043] ;

[0044] in, C is the spatial correlation function, d i is the Euclidean distance from the target node to the corresponding neighboring node, θ i is the local gradient direction of the stress field, is the direction from the target node to the corresponding neighboring node, β is the parameter that controls the decay speed of the direction correlation, ψ Parameters to control the distance attenuation speed;

[0045] Step S54: Calculate weight coefficients based on spatial correlation w i :

[0046] ;

[0047] Standardize the weight coefficient to obtain the standardized weight coefficient w ci :

[0048] ;

[0049] Step S55, calculating the stress of the target node according to the stress of the neighboring nodes:

[0050] ;

[0051] Where S is the stress of the target node, s i is the stress of the neighboring node corresponding to the target node, n l is the number of neighboring nodes corresponding to the target node;

[0052] The stresses of all target nodes form the stress fields of the mesh nodes after meshing.

[0053] In at least one embodiment of the present application,

[0054] In S53, when the grid change is less than the threshold, the spatial correlation function is simplified to:

[0055] ;

[0056] In S54, the weight coefficient is:

[0057] .

[0058] In at least one embodiment of the present application, the method further includes step S56, performing Gaussian filtering or weighted averaging processing on the stress field of the mesh nodes after the mesh is re-meshed.

[0059] In at least one embodiment of the present application, in step S6, the mesh elements of the milled part in the overall stiffness matrix are removed by reorganizing the element stiffness matrix to obtain a part model of the retained part, including:

[0060] Step S61, determining the initial value of the reduction coefficient α, α<1;

[0061] Step S62: Calculate the overall stiffness matrix according to the reduction factor α K :

[0062] ;

[0063] ;

[0064] in, m is the total number of grid cells, n is the number of mesh elements in the milled part, K e is the element stiffness matrix of the component part, is the element stiffness matrix of the milled part;

[0065] Step S63: Calculate the overall stiffness matrix K The condition number cond( K );

[0066] Step S64, obtain the stability judgment threshold ζ, and judge the overall stiffness matrix K The condition number cond( K ) and the stability determination threshold ζ;

[0067] When cond( K )>ζ, increase the reduction coefficient α and return to step S62;

[0068] When cond( K )≤ζ, the overall stiffness matrix that meets the stability requirements is obtained K ;

[0069] Step S65: Determine the reduction coefficient α and the impact determination threshold μ When α> μ When α is reduced, the reduction factor α is reduced, and steps S62-S64 are repeated until α= μ , and the final overall stiffness matrix is ​​obtained K ;

[0070] Step S66: According to the overall stiffness matrix KGet the part model of the retained part.

[0071] In at least one embodiment of the present application, in step S63, the overall stiffness matrix K The condition number cond( K )for:

[0072] ;

[0073] Among them, σ max ( K ) is the overall stiffness matrix K The largest element on the diagonal of the singular matrix Σ obtained by singular value decomposition, σ min ( K ) is the overall stiffness matrix K The smallest element on the diagonal of the singular matrix Σ obtained by singular value decomposition, h max The maximum grid edge length of a grid cell, h min The minimum grid edge length of a grid cell, m is the total number of grid cells, is the dimension of the Euclidean space, 2 η is the order of the partial differential equation to be solved.

[0074] The invention has at least the following beneficial technical effects:

[0075] The part milling process simulation method considering the multi-step stress release process of the present application can realize the deformation simulation of the parts processed by the multi-step stress release process, effectively control the deformation amount of the parts processing, and minimize the manpower and material costs of process optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 is a flow chart of a method for simulating a part milling process taking into account a multi-step stress release process according to an embodiment of the present application;

[0077] Figure 2 is a schematic diagram of geometric re-division of an embodiment of the present application;

[0078] Figure 3 It is a schematic diagram of the changes of the part model during the simulation process of one embodiment of the present application. DETAILED DESCRIPTION

[0079] In order to make the purpose, technical scheme and advantages of the implementation of this application clearer, the technical scheme in the embodiment of this application will be described in more detail below in conjunction with the drawings in the embodiment of this application. In the drawings, the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The described embodiments are part of the embodiments of this application, not all of them. The embodiments described below with reference to the drawings are exemplary and are intended to be used to explain this application, and should not be construed as limitations on this application. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application. The embodiments of this application are described in detail below in conjunction with the drawings.

[0080] In the description of the present application, it should be understood that the terms "center", "longitudinal", "lateral", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the scope of protection of the present application.

[0081] The following is combined with Figures 1 to 3 This application is described in further detail.

[0082] This application provides a method for simulating the milling process of a part considering a multi-step stress release process, such as Figure 1 As shown, the following steps are included:

[0083] Step S1, constructing a part model, and performing geometric division and mesh division on the part model;

[0084] Step S2, obtaining initial residual stress, applying the initial residual stress to the part model, and performing simulation calculation to obtain residual stress and deformation;

[0085] Step S3, dividing the part model into a milled part and a part part by geometric redivision;

[0086] Step S4, re-meshing the milled part and the component part respectively;

[0087] Step S5, mapping the residual stress obtained by simulation calculation to the mesh units of the milled part and the component part after mesh re-division, to obtain the residual stress of the milled part and the component part after mesh re-division;

[0088] Step S6, removing the mesh elements of the milled part in the overall stiffness matrix by reorganizing the element stiffness matrix to obtain a part model of the retained part;

[0089] Step S7, applying the residual stress of the component obtained in step S5 to the component model obtained in step S6, performing simulation calculation to obtain the residual stress and deformation, and returning to step S3;

[0090] Step S8, looping steps S3 to S7 until the part model reaches a predetermined thickness.

[0091] The part milling process simulation method of the present application taking into account the multi-step stress release process, in step S1, first constructs a part model, obtains the material information and clamping boundary conditions of the part, constructs a blank material model of the part as the part model based on the material information and clamping boundary conditions of the part, and further performs preliminary geometric division and mesh division on the part model.

[0092] In the part milling process simulation method considering the multi-step stress release process of the present application, in step S2, the initial residual stress is obtained through experiments, and the initial residual stress is applied to the part model in layers. The residual stress of each layer is obtained by linear interpolation method:

[0093] ;

[0094] ;

[0095] ;

[0096] ;

[0097] in, E e is the residual stress of the corresponding layer, H is the height of the interpolation layer, H k is the height of the k-th floor, H k-1 is the height of the k-1th floor, A is the cross-sectional area, and M is the moment;

[0098] The interpolation layer height is:

[0099] H=Cd

[0100] Where d is the milling depth and C is a positive integer.

[0101] In this embodiment, the residual stress of each layer is obtained by linear interpolation, and the residual stress distribution finally applied should comply with the force balance principle and the moment balance principle. Preferably, the layer height H of the interpolation layer is an integer multiple of the milling depth d of the corresponding layer. During the simulation calculation, the stress release process is realized by stress transfer and boundary condition release, and the part is simulated in a state without external stress.

[0102] The part milling process simulation method considering the multi-step stress release process of the present application requires geometric re-division and mesh re-division of the part model with a certain amount of deformation after stress release is achieved through simulation calculation. The processed structure is divided into two parts, the milled part and the part part after the milled part is removed, through geometric re-division. The subdivision surface during geometric re-division is consistent with the processing plane. In this embodiment, in step S3, the subdivision surface when the part model is geometrically re-divided is determined by the following method:

[0103] h-hs>ha

[0104] Among them, h is the thickness of the part, ha is the maximum deformation of the part, and hs is the distance from the sectioning surface to the bottom of the part.

[0105] The simulation process corresponds to the machining process, and each machining step corresponds to a nonlinear static calculation step. Therefore, each calculation step needs to inherit the deformation and stress distribution obtained in the previous calculation step. The subdivision surface during geometric redivision can correspond to the amount of material removed during the machining process. If the geometry and mesh are not redivisioned, the model surface will be too rough to be calculated after removing some mesh elements.

[0106] In the part milling process simulation method of the present application that considers a multi-step stress release process, in step S5, the residual stress obtained by simulation calculation in the first cycle is the residual stress obtained by simulation calculation in step S2, and in subsequent cycles is the residual stress obtained by simulation calculation in step S7.

[0107] The transmission of stress distribution requires remapping of the stress field due to the change of the grid. Since the geometric deformation and stress distribution obtained in each calculation step cannot be determined in advance, coupled with the influence of grid re-division, stress mapping between multiple calculation steps is difficult. In order to more accurately map stress information between different grids. Preferably, in step S5, stress mapping is realized based on the inverse distance weighted method of spatial correlation. By introducing the spatial correlation function and improving the weight calculation method, the accuracy of stress interpolation is improved. The specific steps are as follows:

[0108] Step S51: Select a reference node on the mesh before mesh re-division p 0 ( x i,y i ,z i );

[0109] Select the target node on the remeshed mesh p ( x,y,z );

[0110] Step S52, determining the influence domain of each target node in the mesh before mesh re-division, wherein a spherical area with the target node as the sphere center and R as the radius is taken as the influence domain;

[0111] Search in the influence domain corresponding to each target node:

[0112] If there is a reference node located at the center of the influence domain, the stress of the reference node located at the center of the influence domain is used as the stress of the target node;

[0113] If there is no reference node located at the center of the influence domain, the reference node inside the influence domain is used as the neighboring node;

[0114] By eliminating the reference nodes located at the edge or outside of the influence domain, the reference nodes located inside the influence domain are retained as neighboring nodes for the next step of spatial correlation calculation;

[0115] Step S53: Calculate the spatial correlation between the target node and the neighboring nodes according to the spatial correlation function:

[0116] ;

[0117] ;

[0118] in, C is the spatial correlation function, d i is the Euclidean distance from the target node to the corresponding neighboring node, θ i is the local gradient direction of the stress field, is the direction from the target node to the corresponding neighboring node, β is the parameter that controls the decay speed of the direction correlation, ψ Parameters to control the distance attenuation speed;

[0119] Step S54: Calculate weight coefficients based on spatial correlation w i :

[0120] ;

[0121] Standardize the weight coefficient to obtain the standardized weight coefficient w ci :

[0122] ;

[0123] Step S55, calculating the stress of the target node according to the stress of the neighboring nodes:

[0124] ;

[0125] Where S is the stress of the target node, s i is the stress of the neighboring node corresponding to the target node, n l is the number of neighboring nodes corresponding to the target node;

[0126] The stresses of all target nodes form the stress fields of the mesh nodes after meshing.

[0127] It can be understood that in step S53, when the grid changes dramatically, the spatial correlation function considers the direction and angle changes between nodes. When the grid changes slightly, for example, less than a certain threshold, the direction and angle related parameters can be approximately equal to zero, and the spatial correlation function is simplified to:

[0128] ;

[0129] In this embodiment, in step S54, the weight coefficient is:

[0130] .

[0131] Since the stress of the target node after mesh re-division may not be smooth in the whole domain, in the preferred embodiment of the present application, step S56 is also included, in which the stress field of the mesh node after mesh re-division is subjected to Gaussian filtering or weighted averaging to obtain a smoother stress field.

[0132] The present application discloses a method for simulating the part milling process taking into account a multi-step stress release process. In step S6, the mesh elements of the milled part are removed by reorganizing the element stiffness matrix, and only the part part is retained in the overall stiffness matrix as a new part model to simulate the material removal process.

[0133] Since a large number of mesh elements may need to be removed at one time in this step, if all these stiffnesses are directly set to 0 in the overall stiffness matrix using the general method, it may lead to instability and non-convergence of the calculation. Therefore, the progressive stiffness reduction algorithm is introduced. According to the condition number cond( K ) Dynamically adjust the progressive reduction coefficient α. Specifically, it includes the following steps:

[0134] Step S61, determining the initial value of the reduction coefficient α, α<1;

[0135] Step S62: Calculate the overall stiffness matrix according to the reduction factor α K :

[0136] ;

[0137] ;

[0138] in, m is the total number of grid cells, n is the number of mesh elements in the milled part, K e is the element stiffness matrix of the component part, is the element stiffness matrix of the milled part;

[0139] Step S63: Calculate the overall stiffness matrix K The condition number cond( K );

[0140] Step S64, obtain the stability judgment threshold ζ, and judge the overall stiffness matrix K The condition number cond( K ) and the stability determination threshold ζ;

[0141] When cond( K )>ζ, increase the reduction coefficient α and return to step S62;

[0142] When cond( K )≤ζ, the overall stiffness matrix that meets the stability requirements is obtained K ;

[0143] Step S65: Determine the reduction coefficient α and the impact determination threshold μ When α> μ When α is reduced, the reduction factor α is reduced, and steps S62-S64 are repeated until α= μ , and the final overall stiffness matrix is ​​obtained K ;

[0144] Step S66: According to the overall stiffness matrix K Get the part model of the retained part.

[0145] In this embodiment, the stability determination threshold ζ is generally 10 3 to 10 6 By looping through steps S62-S64, a good overall stiffness matrix is ​​found. K .

[0146] Overall stiffness matrix K The condition number of the general use of the two-norm condition number, the overall stiffness matrix KThe stability of the solution is evaluated, and the two-norm condition number is:

[0147] ;

[0148] When the condition number cond( K ) is too large, indicating that the calculation may cause instability.

[0149] In this embodiment, in order to avoid the reduction of computational efficiency caused by matrix inversion, in step S63, the condition number is approximately calculated as:

[0150] ;

[0151] Among them, σ max ( K ) is the overall stiffness matrix K The largest element on the diagonal of the singular matrix Σ obtained by singular value decomposition, σ min ( K ) is the overall stiffness matrix K The smallest element on the diagonal of the singular matrix Σ obtained by singular value decomposition, h max The maximum grid edge length of a grid cell, h min The minimum grid edge length of a grid cell, m is the total number of grid cells, is the dimension of the Euclidean space, 2 η is the order of the partial differential equation to be solved.

[0152] In step S65, the relationship between the reduction coefficient α and the overall stiffness matrix is:

[0153] ;

[0154] The impact judgment threshold μ is the minimum value that is not zero, which is generally set to 10 -6 .

[0155] Through steps S61-S65, the overall stiffness matrix corresponding to the minimum reduction factor α that meets the stability requirements can be found K Finally, according to the overall stiffness matrix K Get the part model of the retained part.

[0156] The present application's method for simulating the milling process of a part considering a multi-step stress release process, in step S7, performs simulation calculation on the part model after milling to achieve stress release, and after each execution of stress release, returns to step S3 to re-geometrically re-divide and re-mesh the part model after milling. In step S8, through multiple cycles of steps S3 to S7, the simulation of the multi-step stress release process is achieved until the part model reaches a predetermined thickness, thereby achieving the simulation of the complete process of part milling. By optimizing the removed parts each time during the processing by simulation calculation, the purpose of minimizing the maximum deformation ha of the part after the last removal of the milled part can be achieved, thereby effectively controlling the processing deformation of the part.

[0157] In a specific embodiment of the present application, in step S2, the residual stress distribution obtained by the experiment is applied to the part model in layers, and the layer thickness is 2 mm, which is equal to the milling depth of each layer of 2 mm. Figure 2 As shown, in step S3, the geometric re-division divides the part model into two parts: the milled part and the part part. The dividing surface used is generally a plane. The original thickness of the part is 10 mm. The maximum deformation of the part after stress release in step S2 is 0.5 mm. The distance between the dividing surface and the bottom of the part is 8 mm, so h-hs=2 mm, which is greater than ha=0.5 mm. The changes of the part model during the simulation of this embodiment are shown in FIG. Figure 3 shown.

[0158] The present application's method for simulating the milling process of a part considering a multi-step stress release process constructs a part model, and performs preliminary geometric division and mesh division on the part model; after obtaining the initial residual stress, a simulation calculation is performed to realize stress release; after each execution of stress release, the part model is geometrically and meshed re-divided, and the geometric re-division divides the part model into a milled part and a part part; the mesh units of the milled part are removed by reorganizing the unit stiffness matrix to obtain a new part model; then the part model after milling is simulated and calculated; through multiple cycles, the simulation of the part milling process is realized. The present application fully considers the process of removing the material along the machining plane after stress release deformation through the reorganization of the unit stiffness matrix, and realizes efficient simulation of milling structure deformation; through multiple cycles of simulation calculation, deformation simulation of parts processed by multi-step stress release process is realized, the amount of part processing deformation is effectively controlled, and the manpower and material costs of process optimization are minimized.

[0159] The part milling process simulation method considering the multi-step stress release process of the present application is suitable for the simulation of multi-step stress release processes such as stress-free faceting or unequal height process bosses used to control part processing deformation in actual engineering scenarios. Process optimization based on the results of part deformation simulation can significantly improve the efficiency of process optimization and reduce costs, improve the engineering applicability of such processing technology, and effectively reduce processing deformation. Through key technical means such as geometric re-division, mesh re-division, stress mapping, and unit stiffness matrix reorganization, the simulation of multi-step stress release processes can be better achieved, expanding the special process applicability of the simulation method.

[0160] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for simulating a part milling process considering a multi-step stress release process, characterized in that: include: Step S1, constructing a part model, and performing geometric division and mesh division on the part model; Step S2, obtaining initial residual stress, applying the initial residual stress to the part model, and performing simulation calculation to obtain residual stress and deformation; Step S3, dividing the part model into a milled part and a part part by geometric redivision; Step S4, re-meshing the milled part and the component part respectively; Step S5, mapping the residual stress obtained by simulation calculation to the mesh units of the milled part and the component part after mesh re-division, to obtain the residual stress of the milled part and the component part after mesh re-division; Step S6, removing the mesh elements of the milled part in the overall stiffness matrix by reorganizing the element stiffness matrix to obtain a part model of the retained part; Step S7, applying the residual stress of the component obtained in step S5 to the component model obtained in step S6, performing simulation calculation to obtain the residual stress and deformation, and returning to step S3; Step S8, looping steps S3 to S7 until the part model reaches a predetermined thickness; In step S5, the residual stress obtained by simulation calculation is mapped to the mesh units of the milled part and the part after mesh re-division, so as to obtain the residual stress of the milled part and the part after mesh re-division, including: Step S51: Select a reference node p0 (x i ,y i ,z i ); Select the target node p(x,y,z) on the mesh after mesh re-division; Step S52, determining the influence domain of each target node in the mesh before mesh re-division, wherein a spherical area with the target node as the sphere center and R as the radius is taken as the influence domain; Search in the influence domain corresponding to each target node: If there is a reference node located at the center of the influence domain, the stress of the reference node located at the center of the influence domain is used as the stress of the target node; If there is no reference node located at the center of the influence domain, the reference node inside the influence domain is used as the neighboring node; Step S53: Calculate the spatial correlation between the target node and the neighboring nodes according to the spatial correlation function: Where C is the spatial correlation function, d i is the Euclidean distance from the target node to the corresponding neighboring node, θ i is the local gradient direction of the stress field, φ i is the direction from the target node to the corresponding neighboring node, β is the parameter that controls the decay speed of the direction correlation, and ψ is the parameter that controls the decay speed of the distance; Step S54: Calculate the weight coefficient w according to the spatial correlation i : Standardize the weight coefficient to obtain the standardized weight coefficient w ci : Step S55, calculating the stress of the target node according to the stress of the neighboring nodes: Where S is the stress of the target node, s i is the stress of the neighboring node corresponding to the target node, n l is the number of neighboring nodes corresponding to the target node; The stress of all target nodes forms the stress field of the mesh nodes after mesh re-division; In step S6, the mesh elements of the milled part in the overall stiffness matrix are removed by reorganizing the element stiffness matrix to obtain a part model of the retained part, including: Step S61, determining the initial value of the reduction coefficient α, α<1; Step S62: Calculate the overall stiffness matrix K according to the reduction factor α: Where m is the total number of mesh elements, n is the number of mesh elements in the milled part, and K e is the element stiffness matrix of the component part, is the element stiffness matrix of the milled part; Step S63, calculating the condition number cond(K) of the overall stiffness matrix K; Step S64, obtaining a stability determination threshold ζ, and determining the relationship between the condition number cond(K) of the overall stiffness matrix K and the stability determination threshold ζ; When cond(K)>ζ, increase the reduction coefficient α and return to step S62; When cond(K)≤ζ, the overall stiffness matrix K that meets the stability requirements is obtained; Step S65, determine the relationship between the reduction coefficient α and the impact determination threshold μ, when α>μ, reduce the reduction coefficient α, repeat steps S62-S64 until α=μ, and obtain the final overall stiffness matrix K; Step S66: Obtain the part model of the retained part according to the overall stiffness matrix K.

2. The part milling process simulation method considering a multi-step stress release process according to claim 1 is characterized in that: In step S1, a part model is constructed, including: Obtain material information and clamping boundary conditions of parts; Build part models based on part material information and clamping boundary conditions.

3. The part milling process simulation method considering a multi-step stress release process according to claim 2 is characterized in that: In step S2, initial residual stress is obtained and applied to the part model, including: Obtain initial residual stress through testing; The initial residual stress is applied layer by layer to the part model, where the residual stress of each layer is obtained by linear interpolation method: H≤H≤H ∫EedA=0; ∫EedM=0; Among them, E e is the residual stress of the corresponding layer, H is the layer height of the interpolation layer, Hk is the layer height of the k layer, and H k-1 is the height of the k-1th floor, A is the cross-sectional area, and M is the moment; The interpolation layer height is: H=Cd Where d is the milling depth and C is a positive integer.

4. The part milling process simulation method considering a multi-step stress release process according to claim 3 is characterized in that: In step S3, the subdivision plane when geometrically redivision the part model is determined by: h-hs>ha Among them, h is the thickness of the part, ha is the maximum deformation of the part, and hs is the distance from the sectioning surface to the bottom of the part.

5. The part milling process simulation method considering a multi-step stress release process according to claim 4 is characterized in that: In step S53, when the grid change is less than the threshold, the spatial correlation function is simplified to: In step S54, the weight coefficient is:

6. The method for simulating a part milling process taking into account a multi-step stress release process according to claim 5, characterized in that: The method further includes step S56, performing Gaussian filtering or weighted averaging processing on the stress field of the mesh nodes after the mesh is re-divided.

7. The part milling process simulation method considering a multi-step stress release process according to claim 6 is characterized in that: In step S63, the condition number cond(K) of the overall stiffness matrix K is: Among them, σ max (K) is the largest element on the diagonal of the singular matrix Σ obtained by singular value decomposition of the overall stiffness matrix K, σ min (K) is the minimum element on the diagonal of the singular matrix Σ obtained by singular value decomposition of the overall stiffness matrix K, h max The maximum grid edge length of the grid cell, h min is the minimum grid edge length of the grid unit, m is the total number of grid units, χ is the dimension of the Euclidean space, and 2η is the order of the corresponding partial differential equation to be solved.