Method for predicting mechanical cumulative failure part of transformer winding based on mathematical combination

Abstract

The invention provides a method for predicting a mechanical accumulated failure part of a transformer winding based on mathematical combination, and relates to the technical field of mechanical reliability evaluation and fault prediction of power equipment. The method comprises the following steps: firstly, carrying out a mechanical test on an actual wire cake and a cushion block to obtain a nonlinear constitutive model; secondly, constructing an electricity-magnetism-force coupling finite element model, inputting an actual short-circuit impact current, and simulating deformation quantities of a winding and a cushion block; through iterative updating of residual deformation and a material hardening state, multiple times of cyclic simulation of impact-residual response is completed; and finally, counting the maximum accumulated deformation quantity of each layer of winding, and positioning the position with the deformation increase rate exceeding a threshold value for the first time as a mechanical failure weak area. The method can truly reflect the response rule of a complex structure under the short-circuit working condition, improves the accuracy and credibility of failure prediction, and is suitable for the anti-short-circuit design and residual life evaluation of the transformer.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mechanical reliability evaluation and fault prediction of power equipment, and particularly relates to a transformer winding mechanical cumulative failure site prediction method based on mathematical combination. BACKGROUND

[0002] With the rapid development of modern power systems towards high voltage, large capacity and intelligentization, as the core hub of power conversion and transmission, the operation reliability of transformers has become a key element to guarantee the safety and stability of power grids and economic dispatching. Under this background, the cumulative failure prediction method of transformers as an important technical means has gradually attracted widespread attention and research.

[0003] In the aspect of material mechanical property research, the existing work preliminarily obtains the mechanical parameters such as elastic modulus and yield strength of winding copper material through tensile and bending tests, but there are still two key problems: first, the uniaxial tensile test of traditional straight copper sample cannot accurately characterize the mechanical behavior of winding under short-circuit impact. Second, the preparation process of the test sample may cause deviation of the material mechanical properties. At present, most studies use the copper bar after hammering as the standard sample, which may cause the yield strength of copper material to deviate from the actual situation. The limitations of the above tests make it difficult to accurately characterize the plastic deformation process of winding under multiple short-circuit impacts.

[0004] Further research shows that the cumulative deformation of winding is essentially a cumulative process of residual strain after each short-circuit impact. When the short-circuit electromagnetic force exceeds the yield strength, irreversible plastic deformation of winding occurs, and the residual strain generated by each impact gradually accumulates, leading to irreversible changes in the geometric shape of winding and even damage. At the same time, as a nonlinear elastic support element, the internal micro-crack propagation and energy dissipation caused by interface friction of the cushion block during repeated stress process will weaken the structural constraint ability. The coupling effect of plastic deformation of winding copper material and performance degradation of cushion block makes the local stress distribution of winding continuously change, eventually leading to overall structural instability. The traditional cumulative failure prediction method of transformer ignores the nonlinear viscoelastic relaxation effect of cushion block, resulting in insufficient prediction accuracy of cumulative deformation. SUMMARY

[0005] In order to overcome the deficiencies of the prior art, the purpose of the present application is to provide a transformer winding mechanical cumulative failure site prediction method based on mathematical combination, which solves the existing transformer winding deformation fault prediction problem, especially the cumulative failure and weak area positioning problem in the winding deformation process of transformer.

[0006] To achieve the above purpose, the present application provides the following scheme:

[0007] A transformer winding mechanical cumulative failure site prediction method based on mathematical combination comprises:

[0008] S1, obtaining actual transformer line cake and spacer, processing line cake into line cake with different bending radii and spacer to carry out bending test and compression test respectively, and establishing nonlinear constitutive model of line cake and spacer;

[0009] S2, constructing three-dimensional electro-magnetic-force coupled finite element model of pie-type winding-spacer composite structure according to the constitutive model and structure parameters of the target transformer;

[0010] S3, taking actual short-circuit impulse current as excitation input of the finite element model, and calculating deformation of winding and spacer at the end of the first impulse;

[0011] S4, taking residual deformation and material hardening state obtained in step S3 as initial conditions of the next impulse, and cyclically executing step S3 and step S4 until the deformation increment of any unit exceeds the preset threshold value;

[0012] S5, counting maximum cumulative deformation distribution of each layer of line cake, determining the position of the first deformation increment threshold value exceeding as the mechanical cumulative failure weak area, and outputting the prediction result.

[0013] Preferably, the line cake sample with different bending radii is processed by a cutting machine on the actual transformer single-layer line cake, and the original curvature characteristics and span size parameters of the line cake are completely retained.

[0014] Preferably, the displacement control mode is adopted when the radial bending test is carried out on the line cake sample, and the loading rate is 5mm / min, and the bidirectional stress mode is adopted to cover the inward compression of the medium voltage winding and the outward expansion of the high voltage winding at the same time.

[0015] Preferably, in the three-dimensional electro-magnetic-force coupled finite element model of the pie-type winding-spacer composite structure, each layer of line cake is divided into three layers of grid along the axial direction; 16 spacers are arranged between adjacent line cakes; the constitutive relationship of the spacer material is fitted through the compression test data, and the stress-strain curve of the spacer material has the characteristic of nonlinear stiffness enhancement.

[0016] Preferably, in the short-circuit current impulse simulation, the duration of each impulse is 0.01s, the simulation step is not greater than 1ms, and the extreme value of leakage magnetic flux density and the peak value of winding displacement response are recorded.

[0017] Preferably, the residual deformation and material hardening state are automatically imported into the next simulation through the variable initial geometric grid iteration algorithm, and when the residual deformation is imported based on the updated node coordinates and material hardening parameters, the import error is not greater than 0.5%.

[0018] Preferably, the deformation increment threshold value is 20%.

[0019] Preferably, the geometric mesh iteration algorithm reinitializes the finite element mesh based on the updated node coordinates, and synchronously maps the equivalent plastic strain and the updated yield strength as the initial state of the material for the next impact.

[0020] According to the specific embodiments of the present application, the following technical effects are disclosed.

[0021] Firstly, in order to ensure that the geometric shape of the sample is strictly consistent with the actual operation condition of the transformer, the actual single-layer wire cake of the transformer is processed by a cutting machine, and the original curvature characteristics and span size parameters of the wire cake are completely retained.

[0022] Secondly, the method significantly improves the simulation accuracy and engineering applicability through multi-physical field dynamic coupling modeling. Traditional cumulative deformation prediction methods rely on single mechanical field analysis and ignore the synergistic effect of electromagnetic field and material nonlinearity. The pie-type electric-magnetic-force multi-field coupling model constructed by the present application integrates the nonlinear constitutive relationship of the curved copper and the pad, and represents the magnetic flux density, Lorentz force distribution and deformation under short-circuit impact.

[0023] In addition, the initial geometric mesh iteration algorithm proposed by the present application solves the dynamic simulation problem of cumulative deformation effect under multiple impacts. The residual shape after each impact is automatically updated by the finite element script, and the material hardening parameter is corrected. This method realizes the synchronous iteration of geometric state and material properties for the first time, and can capture the sudden increase of deformation, providing key data support for failure warning.

[0024] Finally, a maximum cumulative deformation- short-circuit number correlation model of each layer winding is established combined with simulation data, and according to the sudden increase rate of deformation, the most failure-prone dangerous area, i.e. the weak layer area, is determined.

[0025] In summary, through the synergistic innovation of multi-physical field coupling, dynamic iteration algorithm, double-index failure judgment and data-driven life model, the present application solves the key technical bottlenecks in material nonlinear representation, cumulative deformation simulation and failure positioning of traditional methods, and provides a high-precision and high-efficiency solution for transformer short-circuit resistance design and intelligent operation and maintenance. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0027] Figure 1A method flowchart provided for the embodiment of the present application is shown in

[0028] Figure 2 A technical route schematic diagram provided for the embodiment of the present application is shown in

[0029] Figure 3 A nonlinear stress-strain curve diagram of winding copper material provided for the embodiment of the present application is shown in

[0030] Figure 4 A stress-strain curve of cushion block material provided for the embodiment of the present application is shown in

[0031] Figure 5 A three-dimensional electric-magnetic-force multi-field coupling finite element model schematic diagram of a split pie type winding-cushion block composite structure constructed based on COMSOL provided for the embodiment of the present application is shown in DETAILED DESCRIPTION

[0032] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the present application.

[0033] The purpose of the present application is to provide a transformer winding mechanical cumulative failure site prediction method based on mathematical combination, which can solve the existing transformer winding deformation fault prediction problem, especially the cumulative failure and weak area positioning problem in the transformer winding deformation process.

[0034] In order to make the above-mentioned purposes, characteristics and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0035] Figure 1 A method flowchart provided for the embodiment of the present application is shown in Figure 1 As shown in the figure, the present application provides a transformer winding mechanical cumulative failure site prediction method based on mathematical combination, which comprises:

[0036] S1, obtaining the actual line pie and cushion block of the target transformer, processing the line pie into line pies with different bending radii and cushion blocks, to respectively carry out bending test and compression test, and establishing a nonlinear constitutive model of the line pie and the cushion block;

[0037] S2, constructing a three-dimensional electric-magnetic-force coupling finite element model of a split pie type winding-cushion block composite structure according to the constitutive model and the structure parameters of the target transformer;

[0038] S3, the actual short-circuit impulse current is taken as the excitation input of the finite element model, and the deformation of the winding and the pad at the end of the first impulse is calculated;

[0039] S4, the residual deformation and material hardening state obtained in step S3 are taken as the initial conditions of the next impulse, and steps S3 and S4 are executed in a loop until the deformation increment of any unit exceeds the preset threshold;

[0040] S5, the maximum cumulative deformation distribution of each layer of wire cake is counted, the position that first exceeds the deformation increment threshold is determined as the mechanical cumulative failure weak area, and the prediction result is output.

[0041] As shown in Figures 2 to 5 The technical route of the embodiment is as follows:

[0042] The actual wire cake and pad of the target model transformer are obtained, the wire cake is processed into wire cakes with different bending radii, and the pad is subjected to mechanical test, and the nonlinear constitutive model data is obtained;

[0043] The parameters of the target model transformer are obtained, and a three-dimensional electro-magnetic-force multi-field coupled finite element model of the pie-type winding-pad composite structure is constructed;

[0044] The actual short-circuit impulse current is taken as the excitation, and the deformation of the winding and the pad is calculated;

[0045] The deformation after the last impulse is retained as the initial state of the next impulse, and the model grid is initialized;

[0046] Steps 3-4 are repeated until the deformation increment of the wire or the pad at a certain position exceeds the threshold;

[0047] The maximum deformation distribution of each layer of wire cake after each short-circuit impulse is sorted, the position that first exceeds the deformation increment threshold is defined as the weak area, and the winding mechanical failure position prediction is completed.

[0048] Figure 3 The copper wire stress-strain curve is shown in the figure. Copper wire is a typical elastic-plastic material. In the initial deformation stage, the deformation of the sample is mainly elastic deformation. With the increase of strain, the stress increases rapidly. The stress-strain curve in this stage changes linearly, and the ratio of the two is the elastic modulus of the material. The bending yield strength of the material under bending load is defined as the bending stress corresponding to the point deviating from the linear point, that is, the position of point B in the figure, then the material enters the plastic deformation state after B point. According to the graphical method, point A is determined as the proportional limit, although the deformation of AB segment is still in the elastic stage, the stress and strain are no longer linearly related. After that, with the continuous increase of stress, the strain growth slows down, showing nonlinear plastic characteristics. At this time, even if the stress value increases slightly or remains unchanged, the plastic deformation is still significant.

[0049] Figure 4 The stress-strain curve of the cushion block under compression test is shown. The cushion block shows significant deformability in the initial loading stage, but as the strain increases, its deformation rate gradually decreases, and finally enters a stable stage, showing a non-linear stiffness enhancement feature.

[0050] Figure 5 The traditional cylindrical high, medium and low voltage winding is reconstructed by using a layered wire pie modeling method, and 16 cushion blocks are arranged between adjacent wire pies. Each turn is divided into 3 layers of grid in the axial direction, and the circumferential grid of the winding-cushion contact area is encrypted to 1mm size. The main grid of the cushion block is hexahedral, and the total number of model grids is about 2.8 million units.

[0051] Preferably, the material mechanics test includes:

[0052] For the cushion block material, the nonlinear constitutive relationship data is obtained by compression test;

[0053] For the winding wire pie, the actual transformer single-layer wire pie is processed by a precision cutting machine, and the multi-bending radius sample of the initial curvature radius and span size is obtained. The radial bending test is carried out, and the stress-strain curve is obtained by taking the arithmetic mean value after five tests at a loading speed of 5mm / min.

[0054] Preferably, the multi-physical field modeling method of the pie-type winding-cushion composite structure includes the following simplification strategies:

[0055] Simplification of the core: The core is made of silicon steel sheets stacked together. The laminated core is used to reduce eddy current loss. According to the transformer drawing, a multi-layer laminated core is first constructed, and then stacked into a whole. The magnetic hysteresis effect in the core is ignored.

[0056] Simplification of the winding: The heights of the high-voltage, medium-voltage and low-voltage windings are consistent. The winding is simplified to a concentric cylindrical structure with the same inner diameter and outer diameter as the actual transformer, and then the winding is further segmented to establish a wire pie model.

[0057] Simplification of the cushion block and the support bar: The cushion block and the support bar, which are stacked by multiple layers of insulation paper board, are simplified to a whole structure with the same thickness. When calculating the leakage magnetic field and the stress of the winding, only the effect of the cushion block is considered, and the supporting effect of the support bar is not considered.

[0058] Simplification of the oil tank: The oil tank is simplified to a rectangular box to constrain the boundary of the internal field of the transformer, so as to avoid the problem of non-convergence of the model. The remaining parts on the inner and outer surfaces of the oil tank are ignored, and the transformer insulating oil, which has no effect on the magnetic field distribution, is not considered.

[0059] Simplification of the clamp: The solid geometric structure of the clamp is ignored, and the compression effect of the clamp is simulated by adding fixed constraints.

[0060] Furthermore, in the multiphysics modeling of transformers with split-type winding structures, the transformer itself has a complex and diverse structure, specifically encompassing components such as support bars, spacers, copper windings, oil tanks, cores, and clamps, which makes the calculation of transformers difficult. Therefore, while retaining the objects to be studied during the modeling process, the model is appropriately simplified. At the same time, considering that the short-circuit time is relatively short, the thermal effects of the windings are ignored.

[0061] Furthermore, in the magnetic-structural coupling simulation of transformers under multiple short circuits, according to the Lorentz force law, the interaction between the short-circuit current and the magnetic field within the winding generates a strong mechanical force. These forces act on the winding structure, potentially causing displacement, deformation, or even damage to the winding. The magnetic field-structural field coupling theory combines these two aspects, enabling a comprehensive reflection of the stress and deformation state of the winding.

[0062] Preferably, in the transformer magnetic-structural coupling simulation under multiple short circuits, the transformer magnetic field can be described by Maxwell's equations, typically a low-frequency electromagnetic field, and the governing equation of the magnetic field is:

[0063]

[0064] In the formula, μ is the magnetic permeability of the material, J is the current density, and A is the magnetic vector potential function.

[0065] The structural field satisfies the following equilibrium equations:

[0066]

[0067] In the formula, σ is the stress tensor, and F is the force density generated by the electromagnetic field.

[0068] The magnetic field and the structural field are coupled through the Lorentz force:

[0069]

[0070] In the formula, F is the electromagnetic force, B is the leakage magnetic flux density, and J is the current density.

[0071] The relationship between strain and stress can be described by the constitutive relation (stress-strain relationship) of linear elastic materials:

[0072]

[0073] In the formula ϵ ij Let ν be the strain tensor, ν be Poisson's ratio, E be Young's modulus, and σ be the strain tensor. ij σ is a component of the stress tensor. kk The stress is the volumetric stress.

[0074] The Von Mises yield criterion is used to determine whether the material yields under the multi-axial stress state:

[0075]

[0076] where σ1, σ2, σ3 are the principal stresses, σ e is the equivalent stress.

[0077] Under the Von Mises yield criterion, the expression of the equivalent plastic strain is as follows:

[0078]

[0079] where ε1, ε2, ε3 are the principal strains; ε e is the equivalent strain.

[0080] Preferably, in the multiple short-circuit impact simulation, the actual short-circuit impact current is used as the excitation, each impact lasts for 0.01 s, and the simulation step is ≤1 ms.

[0081] Specifically, the finite element script is used to automatically update the node coordinates, material hardening parameters and pad contact state (residual deformation error ≤0.5%), and to transfer the cumulative deformation effect.

[0082] Preferably, the variable initial geometric mesh iteration algorithm is used, the residual deformation after each impact is used as the initial condition input for the next simulation, the model geometry state is updated to represent the deformation cumulative effect, and the calculation stability is ensured, thereby solving the dynamic simulation problem of the deformation cumulative effect under multiple impacts, and the specific steps are as follows:

[0083] Initial state setting: set the initial geometric model (undeformed state) and the initial hardening parameters of the material (such as the initial yield strength ).

[0084] Single impact simulation: using the actual short-circuit current (I) as the excitation, the electric-magnetic-force multi-field coupling model is solved in the set time step (Δt) , and the displacement field of the winding at the end of this impact and the displacement field of the pad , and the material internal state variables (such as the equivalent plastic strain , the current yield strength ) are calculated.

[0085] Geometry state update: the displacement field at the end of this impact (u , ) is superimposed on the model node coordinates.

[0086]

[0087] wherein , is the node coordinate before the n-th impact, , is the node coordinate before the n+1-th impact (i.e. after the n-th impact).

[0088] Based on the updated node coordinates , The finite element mesh is reinitialized as the initial geometry configuration for the next (n+1-th) impact simulation. This step is automatically done through the script interface of the finite element software (e.g. COMSOL, ANSYS).

[0089] Material state update: The material state variables (mainly the equivalent plastic strain and the updated yield strength ) calculated at the end of the current impact are mapped to the updated mesh. The updated material parameters etc. are assigned to the corresponding elements of the updated mesh as the initial material state for the n+1-th impact.

[0090] Compared with the traditional method, the geometry mesh iteration algorithm used in the present application not only updates the geometry mesh, but also crucially synchronously updates the material constitutive state, significantly improving the accuracy of cumulative plastic deformation prediction.

[0091] Preferably, the deformation increment threshold is specified as 20%. The weak zone positioning method is to arrange the maximum value distribution of the deformation variable of each layer of the line cake after each short-circuit impact, and define the position first exceeding the deformation increment threshold as the weak zone.

[0092] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between various embodiments can be referred to each other.

[0093] The principles and implementation modes of the present application are described by applying specific examples in this paper. The above description of the embodiments is only to help understand the method of the present application and its core idea; at the same time, for the general technical personnel in the art, according to the idea of the present application, the specific implementation mode and application range will be changed. In view of the above, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A mathematical combination-based transformer winding mechanical cumulative failure site prediction method, characterized in that, The method comprises the following steps: S1, obtaining the actual line cake and the cushion block of a target transformer, processing the line cake into line cakes with different bending radii and the cushion block to respectively carry out bending test and compression test, and establishing a nonlinear constitutive model of the line cake and the cushion block; S2, constructing a three-dimensional electro-magnetic-force coupled finite element model of the pie-type winding-cushion block composite structure according to the constitutive model and the structure parameters of the target transformer; S3, taking the actual short-circuit impulse current as an excitation input of the finite element model, and calculating the deformation amount of the winding and the cushion block at the end of the first impulse; S4, taking the residual deformation and the material hardening state obtained in step S3 as the initial conditions of the next impulse, and cyclically executing steps S3 and S4 until the deformation increment of any unit exceeds a preset threshold value; S5, counting the distribution of the maximum cumulative deformation amount of each layer of line cake, determining the position that first exceeds the deformation increment threshold value as the mechanical cumulative failure weak area, and outputting the prediction result.

2. The mathematically combined transformer winding mechanical cumulative failure site prediction method according to claim 1, characterized by, The line cake sample with different bending radii is processed by a cutting machine on the actual transformer single-layer line cake, and the original curvature characteristics and span size parameters of the line cake are completely retained.

3. The mathematically combined transformer winding mechanical cumulative failure site prediction method according to claim 1, characterized by, When the bending test is performed on the line cake sample in the radial direction, a displacement control mode is adopted, and the loading rate is 5mm / min, and the inward compression in the medium voltage winding and the outward expansion of the high voltage winding are simultaneously covered in a bidirectional stress mode.

4. The mathematically combined transformer winding mechanical cumulative failure site prediction method according to claim 1, characterized by, In the three-dimensional electro-magnetic-force coupled finite element model of the pie-type winding-cushion block composite structure, each layer of line cake is divided into three layers of grids along the axial direction; 16 cushion blocks are arranged between adjacent line cakes; the constitutive relationship of the cushion block material is fitted through the compression test data, and the stress-strain curve thereof has the nonlinear stiffness enhancement characteristic.

5. The mathematical combination-based transformer winding mechanical cumulative failure site prediction method according to claim 1, characterized by, In the short-circuit current impact simulation, the duration of each impact is 0.01s, the simulation step is not greater than 1ms, and the leakage magnetic flux density extreme value and the winding displacement response peak value are recorded.

6. The mathematical combination-based transformer winding mechanical cumulative failure site prediction method according to claim 1, characterized by, The residual deformation and the material hardening state are automatically imported into the next simulation through a variable initial geometric grid iteration algorithm, and when the residual deformation is imported based on the updated node coordinates and the material hardening parameters, the import error is not greater than 0.5%.

7. The mathematically combined transformer winding mechanical cumulative failure site prediction method of claim 1, wherein, The deformation increment threshold value is 20%.

8. The mathematically combined transformer winding mechanical cumulative failure site prediction method of claim 6, wherein, The geometric grid iteration algorithm reinitializes the finite element grid based on the updated node coordinates, and synchronously maps the equivalent plastic strain and the updated yield strength as the initial state of the material in the next impact.

Images