Modeling method for assembly interface contact characteristics of high-end equipment driven by measured data

Abstract

The invention provides an actual measurement data driven high-end equipment assembly interface contact characteristic modeling method, which comprises the following steps: S1, aiming at an assembly interface of an aircraft engine rotor or a liquid rocket turbine pump, obtaining actual measurement assembly data and storing the actual measurement assembly data in a key shape force parameter database; s2, based on the key shape force parameter database, adopting contact mechanics to construct an assembly interface contact characteristic mechanical model, and performing interface contact characteristic mechanism analysis to obtain an influence rule of multiple random factors of an actual structure / process / morphology on the assembly interface contact characteristic; s3, establishing an assembly interface gradient virtual material characterization method fusing test data, and developing a three-dimensional 8-node gradient thin layer unit adaptive to non-uniform contact characteristics based on an ABAQUS software UEL subprogram; and S4, modeling to obtain a mechanism and actual measurement assembly data driven high-end equipment assembly interface contact characteristic model. The method has the beneficial effects that the high-efficiency and high-precision high-end equipment assembly interface contact characteristic calculation requirement can be met.

Description

Technical Field

[0001] This invention relates to the technical field of precision assembly and digital assembly, and more specifically, to a method for modeling the contact characteristics of high-end equipment assembly interfaces driven by measured data. Background Technology

[0002] The aerospace industry is a concentrated reflection of a nation's comprehensive national strength and a strategic field driving national defense, technological innovation, and economic and social development. The assembly process is a crucial stage in the development of aero-engines and rocket engines, largely determining their performance. Currently, while my country's manufacturing capabilities are roughly on par with and nearing their limits compared to foreign counterparts, the performance indicators of aero-engines and rocket engines still fall short of the requirements for similar foreign engines. Therefore, how to implement effective assembly after manufacturing to ensure engines meet performance requirements has become a key bottleneck restricting the high-quality development of next-generation aero-engines.

[0003] Assembly interfaces disrupt structural continuity, making assembly connection performance a key factor in assembly service performance. An assembly interface refers to the connection surface between assembled mechanical parts, components, and sub-assemblies, and is widely present in mechanical equipment assembled from parts. With the increasingly extreme service conditions of aerospace and rocket engines, problems related to assembly interfaces are becoming more frequent. Therefore, in-depth research on the contact characteristics of assembly interfaces is urgently needed.

[0004] Phenomenon 1: The random connection status of the assembly interface of the aero-engine rotor during engineering has led to inaccurate prediction of assembly performance and difficulty in precise control based on actual conditions, thus restricting the high-precision and high-efficiency assembly requirements of the new generation of aero-engine rotors. Therefore, strengthening the digital upgrade of traditional aviation equipment assembly technology, realizing efficient and accurate digital assembly analysis of the new generation of aero-engine rotors, improving rotor assembly connection performance, and narrowing the gap with foreign rotor assembly technology levels have become common challenges faced by industry experts.

[0005] Phenomenon 2: In current engineering projects, the random connection status of the assembly interface of liquid rocket turbopumps leads to uneven contact stress at the assembly interface, causing "leakage" during testing. This restricts the high reliability of the assembly connection performance required for the "recovery and reuse" of liquid rocket engines. Therefore, improving and effectively maintaining the uniformity of contact stress distribution at the critical connection sealing assembly interface of liquid rocket engines during service is crucial to ensuring the engine's connection sealing performance, high reliability and durability, and meeting the requirements for reusability. This has become a formidable challenge for aerospace experts.

[0006] The randomness of manufacturing errors and assembly process parameters presents the following research challenges for the efficient and accurate modeling and analysis of assembly interface contact characteristics: Existing assembly interface contact models are insufficient to meet the demands for efficient and high-precision calculations of the contact characteristics of aero-engine rotor assembly interfaces. Current research lacks consideration of the physical properties and friction of assembly interfaces under actual working conditions, and the structure and performance mechanisms of the assembly interface remain unclear. Assembly, as a typical mating characteristic, is correlated with manufacturing errors and assembly processes. While the influence of morphological factors can be "forwardly calculated" using measured data, the randomness of assembly processes leads to "deviations" in interface contact characteristics, resulting in unclear actual contact states and difficulty in accurate characterization. Furthermore, although finite element analysis models can meet the accuracy requirements for contact characteristic calculations, the calculation process requires modeling and analyzing the entire component / assembly, making the computational efficiency insufficient for modeling the aero-engine rotor interface. Therefore, there is an urgent need to explore an efficient modeling method centered on the joint. There is a lack of accurate modeling and updating mechanisms for interface contact characteristics that are dynamically consistent with measured data. Currently, commonly used assembly interface contact characteristic modeling methods are based on the overall structure, such as the finite element method and the contact element method. However, this overall structure modeling approach results in modeling efficiency and interface contact characteristic calculation accuracy that cannot meet the high-efficiency and accurate modeling requirements of performance-oriented digital assembly for high-end equipment, thus hindering the development process and performance improvement of high-end equipment. Furthermore, the interface connection state during assembly exhibits dynamic and random changes with process and load variations, leading to uneven contact stress at the assembly interface connection state. Therefore, dynamically establishing a contact characteristic analysis model by locally replacing interface contact characteristic attribute parameters and combining them with measured data is crucial to ensuring dynamic consistency between physical and informational data; however, relevant models and methods are currently lacking.

[0007] In summary, considering the randomness of the connection state of the assembly interface in engineering, there is an urgent need to develop a modeling method for the contact characteristics of the assembly interface of high-end equipment, based on measured data, that can meet the requirements of efficient and high-precision calculation of the contact characteristics of the assembly interface of high-end equipment such as aero-engine rotors. Summary of the Invention

[0008] The technical problem to be solved by this invention is how to meet the requirements of efficient and high-precision calculation of the contact characteristics of the assembly interface of high-end equipment. In order to overcome the defects of the above-mentioned existing technology (or related technology), this invention provides a method for modeling the contact characteristics of the assembly interface of high-end equipment driven by measured data.

[0009] This invention provides a method for modeling the contact characteristics of high-end equipment assembly interfaces based on measured data, comprising the following steps: Step S1: For the assembly interface of the aero-engine rotor or liquid rocket turbopump, the measured assembly data of the assembly interface is obtained by sensor measurement and stored in the key force parameter database. Step S2: Based on the key shape and force parameter database, a mechanical model of the contact characteristics of the assembly interface is constructed using contact mechanics. The mechanism analysis of the interface contact characteristics is performed to obtain the influence law of multiple random factors of actual structure / process / morphology on the contact characteristics of the assembly interface. Based on the mechanical model of the contact characteristics, a layer of virtual material with a preset thickness and continuous non-uniform material properties is added to the assembly interface. Step S3: Based on the contact characteristics of the assembly interface, establish a gradient virtual material characterization method for the assembly interface by integrating test data. Use the ABAQUS software UEL subroutine to develop a three-dimensional 8-node gradient thin-layer element that adapts to non-uniform contact characteristics. Use the finite element method to perform dynamic and static characteristic numerical analysis on the three-dimensional 8-node gradient thin-layer element to obtain the overall element stiffness matrix and unbalanced force matrix as the constitutive relation of the virtual material of the assembly interface. Step S4: Based on the three-dimensional 8-node gradient thin-layer unit and test data, a high-end equipment assembly interface contact characteristic model driven by mechanism and measured assembly data is obtained.

[0010] Compared with existing technologies, the measured data-driven modeling method for contact characteristics of high-end equipment assembly interfaces of this invention has the following advantages: In this invention, the acquisition of measured assembly data in step S1 ensures that the model originates from the real physical world, laying a data foundation for high-precision modeling. Contact mechanics modeling and virtual material addition in step S2 introduce physical mechanisms, ensuring the interpretability of the model. Numerical analysis of the dynamic and static properties of the virtual material using the finite element method in step S3 yields the overall element stiffness matrix and unbalanced force matrix, which are then integrated with the measured assembly data to provide virtual material property parameters, achieving high-fidelity physical information fusion. By developing a dedicated three-dimensional 8-node gradient element to represent the complex assembly interface, computational resources are concentrated in the key contact area, significantly improving computational efficiency while maintaining accuracy. This meets the demand for efficient and high-precision analysis in the digital assembly of high-end equipment. Finally, step S4 performs efficient and accurate modeling of the contact characteristics of the high-end equipment assembly interface driven by mechanism and data.

[0011] In one possible implementation, in step S1, the assembly interface is subjected to interface morphology testing, interface pressure testing, interface stiffness testing, preload testing, and vibration signal testing to obtain measured morphology data, measured pressure data, measured stiffness data, measured process data, and measured response data, which are then included in the key force parameter database.

[0012] Compared with existing technologies, the above technical solution can clearly define the types of key parameters to be measured. These key parameters cover the entire process of performance information before, during and after assembly. Comprehensive data collection provides data support for the subsequent construction of accurate models that can reflect complex working conditions and randomness, fundamentally avoiding model distortion caused by missing or incomplete data.

[0013] In one possible implementation, step S2 further includes: A fractal geometric interface morphology characterization model is established based on the key form-force parameter database. The influence of fractal parameters on the surface profile of the assembly interface is revealed through the fractal geometric interface morphology characterization model. The form-force relationship of the assembly interface is established using Hertzian contact theory as a bridge. Based on the form-force relationship, the load and stiffness model of the assembly interface is constructed using probability statistics and fractal theory. The load and stiffness model reveals the variation law of the load and stiffness power function relationship of the assembly interface.

[0014] Compared with existing technologies, the above-mentioned technical solution, by introducing fractal theory, can more scientifically describe and quantify the microscopic random morphological features of machined surfaces, which is difficult to achieve with traditional roughness parameters. Combined with Hertz contact theory and probability statistics, a bridge is built from microscopic morphology to macroscopic mechanical properties, revealing the essential law of its inherent power-law relationship. Subsequent integration of measured data and fractal theory for mechanism analysis significantly improves the scientific nature and predictive accuracy of the model.

[0015] In one possible implementation, step S2 further includes: Based on the key force parameter database and Chebyshev polynomials, an uncertainty interval model for surface morphology parameters is constructed. This model reveals the variation law of the power law function of stiffness load and the pressure distribution law of the assembly interface under conventional and random factors.

[0016] Compared with existing technologies, the above-mentioned technical solution constructs a model of the uncertainty range of surface morphology parameters using Chebyshev polynomials. For the first time, it quantitatively incorporates random factors such as unavoidable dimensional errors and assembly force fluctuations in actual production into the modeling scope. This enables the model to predict the possible range of changes in contact characteristics, rather than just an ideal value, thus providing a more reliable and practical decision-making basis for evaluating assembly quality and controlling assembly risks.

[0017] In one possible implementation, in step S3, for any point on the virtual material, a local coordinate system for that point and a global coordinate system for the virtual material are established respectively. The local coordinates of the point in the local coordinate system and the global coordinates in the global coordinate system are obtained, and the node coordinates of the node in the global coordinate system are obtained. A transformation relationship between the local coordinates and the global coordinates is established.

[0018] Compared with existing technologies, the above technical solution establishes a transformation relationship between the local coordinate system and the global coordinate system, which provides a prerequisite for subsequent precise mathematical operations such as isoparametric transformation and Gaussian integration, thus ensuring numerical stability and computational accuracy in finite element analysis.

[0019] In one possible implementation, step S3, the step of analyzing and obtaining the overall element stiffness matrix, includes: Step A1: Based on the displacement vector of any point on the virtual material, a matrix-form shape function is obtained, and based on the shape function and the transformation relationship between the local coordinates and the global coordinates, an isoparametric transformation operation is performed to obtain the element interpolation shape function; Step A2: Based on the strain tensor of any point on the virtual material, a differential operator is introduced, and the strain-displacement relationship matrix is ​​obtained based on the differential operator and the element difference shape function; Step A3: Based on the stress vector of the strain tensor at any point on the virtual material, introduce the property elastic matrix of the virtual material, and obtain the stress-strain relationship matrix according to the stress vector and the property elastic matrix; Step A4: Obtain the overall unit stiffness matrix based on the unit volume of the virtual material, the strain-displacement relationship matrix, and the stress-strain relationship matrix.

[0020] Compared with existing technologies, the above technical solution decomposes the complex finite element development process into four clear and programmable steps, which calculate the shape function, strain-displacement relationship matrix, stress-strain relationship matrix and overall element stiffness matrix respectively, providing a clear and implementable element development algorithm flow.

[0021] In one possible implementation, after performing step S3, the method further includes: The thin-layer unit program for the virtual material is developed using the ABAQUS software subroutine UEL and associated with the ABAQUS software kernel interface for use.

[0022] Compared with existing technologies, the above technical solution enables the proposed high-end equipment assembly interface contact characteristic modeling method to be directly embedded into existing and widely recognized engineering simulation workflows. This greatly reduces the learning and application threshold of new technologies, which is conducive to the rapid promotion and engineering application of the high-end equipment assembly interface contact characteristic modeling method, transforming it from a theoretical method into a practical engineering tool.

[0023] In one possible implementation, in step S3, the developed three-dimensional 8-node gradient element is a ULE element with non-uniformly distributed interface stiffness and replaceable attribute parameters.

[0024] In one possible implementation, step S4 includes: A mechanism- and data-driven assembly interface fusion model framework based on UEL and intelligent algorithms is proposed, which includes physical space and digital space. The physical space mainly includes real-time data acquisition and data storage of the assembly, while the digital space mainly includes assembly interface modeling, data expansion, AI training, AI data fusion, AI discovery, AI inference, fusion model update, and fusion model verification. A mechanism- and data-driven assembly interface contact characteristic dynamic fusion model and error compensation mechanism are also established.

[0025] Compared with existing technologies, the assembly interface contact characteristic model established using the above technical solution ensures the high precision and high efficiency of the high-end equipment assembly interface contact characteristic model, and provides technical guidance for real-time calculation and analysis of interface contact characteristics. Attached Figure Description

[0026] Figure 1 This is a flowchart of the steps of the present invention; Figure 2 This is a schematic diagram of the test data acquisition process of the present invention; Figure 3 This is a schematic diagram of the mechanism modeling of the present invention; Figure 4 This is a schematic diagram illustrating the development of the three-dimensional 8-node gradient unit of the present invention; Figure 5 This is a schematic diagram of the overall framework of the present invention. Detailed Implementation

[0027] First, those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention. Those skilled in the art can make adjustments as needed to adapt to specific application scenarios.

[0028] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0029] See Figure 1-Figure 5 This invention discloses a method for modeling the contact characteristics of high-end equipment assembly interfaces driven by measured data, comprising: Step S1: For the assembly interface of the aero-engine rotor or liquid rocket turbopump, the measured assembly data of the assembly interface is obtained by sensor measurement and stored in the key force parameter database. Step S2: Based on the key shape and force parameter database, a mechanical model of the contact characteristics of the assembly interface is constructed using contact mechanics. The mechanism analysis of the interface contact characteristics is performed to obtain the influence law of multiple random factors of actual structure / process / morphology on the contact characteristics of the assembly interface. Based on the mechanical model of contact characteristics, a layer of virtual material with a preset thickness and continuous non-uniform material properties is added to the assembly interface. Step S3: Based on the contact characteristics of the assembly interface, establish a gradient virtual material characterization method for the assembly interface by integrating test data. Use the UEL subroutine of ABAQUS software to develop a three-dimensional 8-node gradient thin-layer element that adapts to non-uniform contact characteristics. Use the finite element method to perform dynamic and static characteristic numerical analysis on the three-dimensional 8-node gradient thin-layer element to obtain the overall element stiffness matrix and unbalanced force matrix as the constitutive relation of the virtual material of the assembly interface. Step S4: Based on the three-dimensional 8-node gradient thin-layer unit and test data modeling, a high-end equipment assembly interface contact characteristic model driven by mechanism and measured assembly data is obtained.

[0030] In this embodiment of the invention, see Figure 2 Considering the typical randomness of the aero-engine rotor assembly process, and the lack of analysis of interface contact characteristics under the influence of random factors in existing research, the analysis results differ from the physical reality. Therefore, in step S1, a key form / force parameter testing process for the assembly interface contact characteristics considering the actual structure / morphology / process of the aero-engine rotor is first established. This includes interface morphology testing, interface pressure testing, interface stiffness testing, preload testing, and vibration signal testing to obtain measured morphology data, pressure data, stiffness data, process data, and response data, thereby acquiring measured manufacturing / assembly errors and assembly performance data. For the instability, missing points, and noise in the above measured assembly data, mathematical methods such as data mapping, fitting algorithms, and filtering budgets are used to correct the data before storing it in the key form / force parameter database.

[0031] In this embodiment of the invention, see Figure 3Based on the obtained measured assembly data, a mechanical model of the contact characteristics of the assembly interface is established using contact mechanics: a fractal geometric interface morphology characterization model is established to reveal the influence of fractal parameters on the surface profile; Hertzian contact theory is used as a "bridge" to establish the form / force relationship of the assembly interface; a deformation mechanical model of the assembly interface is constructed; and then, a load and stiffness model of the assembly interface is constructed using probability statistics and fractal theory to reveal the variation law of the relationship between the load and stiffness power function of the assembly interface. Taking into account multiple random factors, from the perspective of macro-micro multi-scale modeling, the nonlinear correlation law between the actual structure / morphology / process of the rotor and the performance of the assembly interface is studied.

[0032] In this embodiment of the invention, an uncertainty interval model of surface morphology parameters is established based on measured assembly data and Chebyshev polynomials. This model reveals the variation law of stiffness load power law function and pressure distribution law of assembly interface under conventional and random factors. Furthermore, considering the dynamic and random characteristics of the process, and given that Bayesian theory has a rigorous probabilistic framework, it can reasonably quantify parameters and model uncertainties by integrating experimental data with physics-based theoretical models. It has high computational efficiency and is widely used in uncertainty analysis, model correction and other research.

[0033] In this embodiment of the invention, see Figure 4 To address the issue of insufficient modeling accuracy in the equivalent model of bolted joints, this paper comprehensively considers the non-uniform pressure distribution at the assembly interface and the power-law nonlinear relationship of stiffness load. By adding a layer of virtual material with a certain preset thickness and continuously non-uniformly distributed material properties between the assembly interfaces, this virtual material is used to equivalently replace the assembly interface model. Based on the principles of material mechanics, elasticity, and interface mechanics, the elastic modulus, shear modulus, Poisson's ratio, density, and other material property parameters of the virtual material layer are accurately defined and calculated according to the principle of equal strain energy.

[0034] In this embodiment of the invention, the finite element method and its basic steps are summarized as follows for the ABAQUS software UEL subroutine development process: Discretization and selection of cell type; Select displacement function N ; Define strain / displacement B and stress / strain relationship D ; Derivation of element stiffness matrix K and equilibrium equations; The overall equation is derived by assembling the unit equations, and boundary conditions are then introduced. Solve for unknown degrees of freedom (or generalized displacements); Solve for element strain and stress; Explain the results; The steps have been completed in the modeling process using ABAQUS software. During the solution process, the ABAQUS software will automatically complete the steps. , , , Therefore, the algorithm steps in this invention focus on calculating the element stiffness matrix and equilibrium equations, i.e., steps... , , In other words, writing UEL subroutines, combined with the UEL subroutine interface, the general algorithm steps for nonlinear analysis of virtual materials at the assembly interface can be summarized as follows: Calculate shape functions N ; calculate B The matrix, i.e., the strain-displacement relationship matrix; calculate D The matrix, i.e., the stress-strain relationship matrix; Calculate the overall element stiffness matrix K ( AMATRX ); Calculate the unbalanced force matrix R ( RHS ); In the specific implementation process, for any point on the virtual material, a local coordinate system and a global coordinate system of the virtual material are established for that point. The local coordinates of the point in the local coordinate system and the global coordinates in the global coordinate system are obtained. The node coordinates of the nodes in the global coordinate system are also obtained. The transformation relationship between local coordinates and global coordinates, as well as the transformation relationship between local coordinates and node coordinates, are established. The steps for analyzing and obtaining the overall element stiffness matrix are detailed as follows: Step A1: Based on the displacement vector of any point on the virtual material, a matrix-form shape function is obtained. Based on the shape function and the transformation relationship between local and global coordinates, an isoparametric transformation operation is performed to obtain the element interpolation shape function. Step A2: Based on the strain tensor at any point on the virtual material, a differential operator is introduced, and the strain-displacement relationship matrix is ​​obtained based on the differential operator and the element difference shape function. Step A3: Based on the stress vector of the strain tensor at any point on the virtual material, introduce the property elastic matrix of the virtual material, and obtain the stress-strain relationship matrix according to the stress vector and the property elastic matrix. Step A4: Obtain the overall element stiffness matrix based on the element volume, strain-displacement matrix, and stress-strain matrix of the virtual material; Then, the virtual work done by the external load when the virtual material is subjected to static load is analyzed to obtain the concentrated force, surface force and body force of the external load acting on the virtual material, and the unbalanced force matrix is ​​constructed based on the shape function, concentrated force, surface force and body force.

[0035] In this embodiment of the invention, Calculate shape functions N The calculation process can be broken down as follows: Express the displacement function in matrix form: (1) In the formula, Let be the displacement vector of any point in a 3D 8-node gradient element, i.e. ; N For shape functions, ,in: (2) is the displacement vector of the element node; Perform isoparametric transformation operations, element interpolation shape function Since it is a function of local coordinates, we need to establish the relationship between the shape function and its derivatives with respect to the global and local coordinates. First, we take the partial derivative with respect to the shape function, and obtain: (3) Equation (3) is then written in matrix form: (4) Define the Jacobian matrix J for: (5) The local coordinates of any point in a 3D 8-node gradient element. With node coordinates Relationship Substituting into equation (5), we get: (6) Due to the unit interpolation shape function Local coordinates x , h、z The function is such that its partial derivative with respect to local coordinates is: (7) Therefore, knowing the local coordinates of any point in the three-dimensional 8-node gradient unit and the global and local coordinates of each node, the Jacobian matrix can be obtained by equation (6). In addition, from equation (4), we can obtain: (8) in, Given the inverse of the Jacobian matrix, further extending equation (8) yields: (9) The Jacobian matrix was calculated. J After obtaining the partial derivatives of the shape functions, we can calculate the result using equation (9). strain matrix B Can be derived from matrix We obtain it by performing the corresponding formal transformations.

[0036] In this embodiment of the invention, calculate B The matrix, i.e., the strain-displacement relationship matrix, can be refined into the following calculation process: According to the geometric equations of elasticity: (10) In the formula, Let be the strain tensor at any point within a 3D 8-node gradient element. , L For the differential operator, the expression is as follows: (11) Substituting equation (1) into equation (10), we get: (12) In the formula, (13) (14) (15).

[0037] In this embodiment of the invention, calculate D The matrix, i.e., the stress-strain relationship matrix, can be refined into the following calculation process: According to the physical equations of elasticity, we know that: (16) In the formula, Let be the stress vector of the strain tensor at any point within a three-dimensional 8-node gradient element. ; D It is an elastic matrix related to material properties.

[0038] In this embodiment of the invention, Calculate the overall element stiffness matrix K The calculation process can be broken down as follows: (17) For any given element, the element volume is... V The unit area is S Suppose that at a certain moment, the virtual displacement vector of each node of the element is... The virtual displacement vector of a point within an element Then the virtual strain corresponding to the virtual displacement is: (18) Therefore, the virtual strain energy within the element can be expressed as: (19) When the element is subjected to a static load, the virtual work done by the applied load can be expressed as: (20) When the element is subjected to dynamic load, it is simultaneously subjected to inertial force caused by acceleration. Damping force caused by velocity Functions, among which For material density, c Let be the linear damping coefficient of the material. In this case, the virtual work done by the external force can be expressed as: (twenty one) In equations (20) and (21), , , These are concentrated forces, surface forces, and volume forces acting on the element, respectively. In addition, shape functions N Coordinates only x , y , z The function is independent of time, therefore we have: (twenty two) (twenty three) (twenty four) (25) According to the principle of virtual work, the work done by external force on virtual displacement is equal to the work done by stress on virtual strain, that is: (26) Substituting equations (18)-(25) into equation (26), we obtain the equilibrium equations for the element under static load: (27) The equation of motion of the element under dynamic load is: (28) The following relationship exists between equations (27) and (28): (29) (30) (31) In the formula, , , These are respectively called the element mass matrix, element stiffness matrix, and element damping matrix. They are the characteristic matrices of the custom element and together determine the static and dynamic performance of the custom element. When performing isoparametric transformation operations, it can be seen from equations (29), (30), and (31) that when calculating the characteristic matrix of a unit using calculus, the volume element is always represented in the global coordinate system. dV Integral calculations require considering the infinitesimal volume element on the local coordinate system. dξdηdζ Integrate points; Infinite element vectors in three directions in the local coordinate system dx , dη、dζ All in x , h、z The one selected above, so dx , dη、dζ On x , y、z They are x , h、z The function can be expressed in the global coordinate system as: (32) A volume element can be represented by a vector product as follows: (33) Substituting equation (32) into equation (33), we get: (34) The calculus operation of the element characteristic matrix in the local coordinate system can be further expressed as: (35) (36) (37) Then, Gaussian integration is performed. According to equations (35), (36), and (37), the element stiffness matrix, damping matrix, and mass matrix are known in the three dimensions of the local coordinate system. x , or , gIntegrating on (-1, 1), according to the Gaussian integral principle, two integration points are chosen on each dimension. , Since the weighting coefficients for both points are 1, we obtain the result consisting of 8 integration points in the local coordinate system. Gauss The integral point matrix is ​​as follows: (38) Therefore, the expressions for the element characteristic matrix and the corresponding numerical integral are: (39) (40) (41) In the formula, , , Here are the weighting coefficients for each integration point of the Gaussian integral. , , All are 1.

[0039] In this embodiment of the invention, Calculate the unbalanced force matrix R The calculation process can be broken down as follows: (42) In the formula, It is the dynamic load vector of the element node, which is obtained by shifting the concentrated force, dynamic surface force and dynamic body force acting on the element to the element node.

[0040] In this embodiment of the invention, after executing step S3, the thin-layer unit program of the virtual material can be compiled using Fortran language based on the ABAQUS software subroutine UEL and associated with the ABAQUS software kernel interface, thereby completing the subroutine UEL call and program debugging for use.

[0041] In this embodiment of the invention, see Figure 5 Furthermore, it can incorporate the concepts of AI-enabled mechanisms and data-driven approaches. Addressing the issues of weak integration and innovation of AI technology in digital assembly, difficulties in data acquisition and processing, and insufficient interaction, fusion, and iterative updates between mechanism models and measured data, this paper proposes an assembly interface fusion model framework based on subroutine UEL, AI-enabled mechanisms, and data-driven approaches. This framework includes physical and digital spaces. The physical space mainly includes real-time data acquisition and storage of the assembly, while the digital space mainly includes assembly interface modeling, data expansion, AI training, AI data fusion, AI discovery, AI inference, fusion model updates, and fusion model verification. This framework provides an overall architecture for constructing a fusion model of assembly interface contact characteristics.

[0042] In this embodiment of the invention, an assembly interface contact characteristic dynamic fusion model and error compensation mechanism jointly driven by AI empowerment mechanism and data can also be established. First, addressing the issues of assembly randomness and the discrepancy between the equivalent digital model and the actual physical model of the joint, an accurate assembly interface model integrating measured assembly data and mechanism model is established based on measured assembly data and three-dimensional 8-node gradient units. To achieve dynamic consistency of the model, a forward given model of virtual materials based on fractal theory and Hertzian contact theory, a reverse given model of virtual materials based on the idea of ​​minimizing dynamic error optimization, and an approximate model given model of virtual material parameters based on variable fidelity proxy model are established respectively. Second, based on three-dimensional 8-node gradient units... The replaceability of meta-parameters is achieved by correcting the parameters of the 3D 8-node gradient elements using measured assembly data, thereby updating the assembly fusion model. Parameters such as the elastic modulus and shear modulus of the elements are optimized and identified using AI-enabled technology, surrogate models, and improved genetic algorithms. This represents a breakthrough in the multidisciplinary integration and innovation of AI algorithms such as deep learning and knowledge graphs in the field of performance-oriented digital assembly research. Finally, the corrected element attribute parameters are fed back to the interface physical / information fusion model to complete the fusion model update, achieving high-performance intelligent control. Each model is validated using dynamic experiments, and the optimal model is selected while considering accuracy and efficiency requirements. This provides a model foundation for subsequent quantitative evaluation of the assembly interface connection state and consistent process optimization control.

[0043] In this embodiment of the invention, step S4 proposes an assembly interface fusion model framework based on UEL and intelligent algorithm empowerment, driven by both mechanism and data. This framework includes a physical space and a digital space. The physical space mainly includes real-time data acquisition and storage of the assembly, while the digital space mainly includes assembly interface modeling, data expansion, AI training, AI data fusion, AI discovery, AI inference, fusion model updating, and fusion model verification. Furthermore, a mechanism- and data-driven assembly interface contact characteristic dynamic fusion model and error compensation mechanism are established. Compared with existing technologies, the assembly interface contact characteristic model established using the above technical solution ensures the high precision and high efficiency of the high-end equipment assembly interface contact characteristic model, providing technical guidance for real-time calculation and analysis of interface contact characteristics.

[0044] In the description of this invention, the references to "one embodiment," "some embodiments," "in this embodiment," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0045] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A measured data driven high end equipment assembly interface contact property modeling method, characterized in that, The method comprises the following steps: Step S1, for the assembly interface of aero-engine rotor or liquid rocket turbine pump, measured assembly data of the assembly interface is obtained by sensor measurement and stored in a key shape force parameter database; Step S2, based on the key shape force parameter database, a contact mechanics is used to construct an assembly interface contact characteristic mechanics model, a mechanism analysis of the interface contact characteristics is performed to obtain the influence law of actual structure / technology / topography multi-randomness factors on the assembly interface contact characteristics, and a virtual material with a preset thickness and continuous and uneven material properties is added to the assembly interface based on the contact characteristic mechanics model; Step S3, based on the assembly interface contact characteristic law, a gradient virtual material characterization method of the assembly interface is established, a three-dimensional 8-node gradient thin layer element suitable for non-uniform contact characteristics is developed by using an ABAQUS software UEL subroutine, and numerical analysis of dynamic and static characteristics of the three-dimensional 8-node gradient thin layer element is performed by using a finite element method to obtain an overall element stiffness matrix and an unbalanced force matrix as a constitutive relation of the virtual material of the assembly interface; Step S4, based on the three-dimensional 8-node gradient thin layer element and the test data, a mechanism and measured assembly data driven high-end equipment assembly interface contact characteristic model is obtained.

2. The high-end equipment assembly interface contact property modeling method of claim 1, wherein, In the step S1, the interface topography test, interface pressure test, interface stiffness test, pre-tightening force test and vibration signal test are performed on the assembly interface to obtain topography measured data, pressure measured data, stiffness measured data, process measured data and response measured data, which are contained in the key shape force parameter database.

3. The measured data driven high end equipment assembly interface contact property modeling method of claim 1, wherein, In the step S2, the following steps are further included: Based on the key shape force parameter database, a fractal geometric interface topography characterization model is established, the influence law of fractal parameters on the surface profile of the assembly interface is revealed through the fractal geometric interface topography characterization model, a shape-force correlation of the assembly interface is established based on the Hertz contact theory, and based on the shape-force correlation, a load and stiffness model of the assembly interface is constructed by using probability statistics and fractal theory, and the power function relationship change law of the load and stiffness of the assembly interface is revealed through the load and stiffness model.

4. The measured data driven high end equipment assembly interface contact property modeling method of claim 1, wherein, In the step S2, the following steps are further included: Based on the key shape force parameter database and Chebyshev polynomials, a surface topography parameter uncertainty interval model is constructed, and through the surface topography parameter uncertainty interval model, the power law function change law of the stiffness and load of the assembly interface and the pressure distribution law under conventional factors and random factors are revealed.

5. The measured data driven high end equipment assembly interface contact property modeling method of claim 1, wherein, In the step S3, for any point on the virtual material, a local coordinate system of the point and a global coordinate system of the virtual material are established, the local coordinate of the point in the local coordinate system and the global coordinate of the point in the global coordinate system are obtained, the node coordinates of the nodes in the global coordinate system are obtained, and the conversion relationship between the local coordinate and the global coordinate is established.

6. The measured data driven high end equipment assembly interface contact property modeling method of claim 5, wherein, In the step S3, the step of analyzing the overall element stiffness matrix comprises: Step A1, based on the displacement vector of the arbitrary point on the virtual material, the matrix form of the shape function is obtained, and based on the shape function and the conversion relationship between the local coordinates and the global coordinates, the isoparametric transformation operation is carried out to obtain the element interpolation shape function; Step A2, based on the strain tensor of the arbitrary point on the virtual material, the differential operator is introduced, and based on the differential operator and the element difference shape function, the strain-displacement relationship matrix is obtained; Step A3, based on the stress vector of the strain tensor of the arbitrary point on the virtual material, the attribute elastic matrix of the virtual material is introduced, and the stress-strain relationship matrix is obtained according to the stress vector and the attribute elastic matrix; Step A4, according to the element volume of the virtual material, the strain-displacement relationship matrix and the stress-strain relationship matrix, the total element stiffness matrix is obtained.

7. The measured data driven high end equipment assembly interface contact property modeling method of claim 1, wherein, After the step S3 is executed, it further includes: Based on the ABAQUS software subroutine UEL, the thin layer element program of the virtual material is programmed by Fortran language, and the ABAQUS software kernel interface is associated for use.

8. The measured data driven high end equipment assembly interface contact property modeling method of claim 1, wherein, In the step S3, the developed three-dimensional 8-node gradient element is a ULE element with non-uniform interface stiffness and replaceable attribute parameters.

Images