Method for improving nonuniformity of magnetic field in magnetic disk space

Abstract

The invention relates to the cross technical field of computational electromagnetism and precision permanent magnet device design, in particular to a method for improving the nonuniformity of a magnetic disk space magnetic field, and aims to solve the problems that a method for improving the uniformity of a magnetic field in the prior art is low in optimization efficiency and difficult to break through the uniformity bottleneck caused by angular harmonic waves and a structural coupling effect. According to the scheme, the method comprises the steps that an accurate finite element model containing seven circles of fan-shaped magnets, a parallel magnetic disk, an eddy current disk and an iron support is established in Opera3D, and meshing and solving are carried out; in the first stage, the structural height of the iron stand is optimized through parameterized scanning, and the optimal height parameter is determined; and in the second stage, on the basis of the optimal iron stand height, the outermost two circles of the magnetic disk are set as NdFeB. The non-uniformity deviation (delta B) of the magnetic field of the target spherical area can be reduced by 13%, and the method is suitable for high-precision scenes such as nuclear magnetic resonance and synchrotron radiation light sources and has the advantages of being efficient in design and high in repeatability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer electromagnetics and precision permanent magnet device design, and particularly relates to a method for improving magnetic field non-uniformity in a disk space. BACKGROUND

[0002] In high-precision electromagnetic applications such as nuclear magnetic resonance (NMR), synchrotron radiation sources, and high-sensitivity magnetic sensors, it is necessary to generate a highly uniform static magnetic field (the non-uniformity of which often needs to be less than several parts per million) in a specific spatial region. An axial magnetization parallel disk system composed of multiple turns of discrete sector-shaped permanent magnet units is one of the mainstream technical solutions for generating such strong magnetic fields. However, the magnetic field uniformity of this structure is subject to multiple complex factors: first, the nonlinear characteristics (such as demagnetization curves) and discrete distribution (sector splicing) of the permanent magnet material will introduce inherent spatial magnetic field harmonics, especially low-order (such as quadrupole and sextupole) angular harmonics, resulting in periodic fluctuations of the magnetic field in the angular direction; second, there is a strong electromagnetic coupling effect between the various components (such as the magnets, magnetically conductive vortex disks, and ferromagnetic support frames) in the system, and the saturation effect of ferromagnetic materials (such as DT4) will nonlinearly change the magnetic flux path; finally, geometric size mutations (such as different heights of the magnets in each turn) will cause dramatic changes in the magnetic flux density gradient.

[0003] The existing technology for improving magnetic field uniformity mainly adopts two methods: one is "passive shimming" after the assembly of the magnets, that is, local compensation is performed by pasting soft magnetic material sheets in the areas where the magnetic field is not uniform, which relies on experience and has limited adjustment freedom; the other is "active shimming", that is, additional controllable current coils are introduced to generate a compensation magnetic field, which is accurate but increases the complexity and energy consumption of the system. In the design stage, computer simulation optimization based on the finite element method has become a forward-looking solution. However, current optimization strategies are mostly limited to adjusting the parameters of the permanent magnets themselves, such as optimizing the radial size of the magnets or implementing radial material gradient distribution, while two key dimensions are generally ignored: the macroscopic electromagnetic shaping effect of the ferromagnetic support structure (such as the iron stand), and the size change of which will significantly affect the magnetic resistance distribution and leakage of the entire magnetic circuit; the suppression means for angular magnetic field harmonics, and the conventional radial optimization has limited effect on eliminating angular non-uniformity. Therefore, the existing design methods often have low optimization efficiency and are difficult to break through the uniformity bottleneck caused by angular harmonics and structural coupling effects. Therefore, the present application proposes a method for improving the magnetic field non-uniformity in a disk space. SUMMARY

[0004] The method for improving the magnetic field non-uniformity in a disk space proposed by the present application solves the problem of low optimization efficiency of the method for improving the magnetic field uniformity in the prior art, and the difficulty in breaking through the uniformity bottleneck caused by angular harmonics and structural coupling effects.

[0005] In order to achieve the above object, the present application adopts the following technical scheme: A method for improving the magnetic field non-uniformity of a magnetic disk space, comprising the following steps: S1, geometric modeling and import: based on a three-dimensional modeling software, a precise model of a magnetic disk system including seven circle sector magnets, parallel magnetic disks, eddy current disks and an iron stand is constructed, and the model is imported into Opera3D finite element analysis software in STEP or IGES format; S2, model attribute definition and solution setting: in Opera3D, the material properties (the iron stand and the eddy current disk are associated with the DT4 material B-H curve, the magnet is associated with the residual magnetization Br and coercive force Hc data of NdFeB / SmCo), the magnetization direction (the magnet is set to axial magnetization) and the boundary condition (based on the symmetry of the model, the 1 / 4 symmetry plane boundary is set) of the iron stand, the eddy current disk and each circle magnet are defined, and the TOSCA Magnetic module is selected for static magnetic field solution; S3, meshing and solution: a cylindrical air domain wrapping the model is created, surface triangular mesh and three-dimensional tetrahedral mesh (tolerance is set to 1.0E-08, and the magnet and air gap area are encrypted mesh) are generated, and solution calculation is performed to obtain the system magnetic field distribution; S4, post-processing analysis: dense sampling points are created in the target spherical region (radius 10 cm) at the geometric center of the two magnetic disks, the magnetic induction intensity Bz component is extracted, and the magnetic field non-uniformity deviation ΔB=(B_max-B_min) / B average is calculated as the non-uniformity index; S5, two-stage system optimization: S51, structure parameter optimization stage: taking reducing the magnetic field non-uniformity of the target region as the target, the height parameter of the iron stand is optimized, and the optimal value is determined through iterative simulation; S52, material distribution optimization stage: on the basis of the optimal iron stand height, the angular and radial distribution ratio of NdFeB and SmCo materials in the seven circle magnetic disk magnets is optimized, and the optimal distribution is determined through iterative simulation.

[0006] Through the above technical scheme, through the collaborative optimization of structure parameters and material distribution, the peak-to-peak value deviation of the target spherical region can be significantly reduced, and an efficient and reliable design means is provided for solving the magnetic field uniformity problem of complex permanent magnet systems.

[0007] As a further improvement of the above scheme, in step S1, the three-dimensional modeling software is SolidWorks, and the radius, sector angle and height (from outside to inside, 4.7 cm, 2.2 cm, 1.95 cm, 1.95 cm, 2.18 cm, 2.24 cm, 2.06 cm) of each circle sector magnet need to be defined separately during modeling, and the model iron stand height is set to 30 cm.

[0008] Through the technical scheme, the accuracy and reproducibility of the modeling process are ensured by limiting the modeling tool and the core geometric parameters, and a unified geometric reference is provided for subsequent simulation and optimization.

[0009] As a further improvement of the above scheme, in step S2, the PotentialType of the iron stand and the eddy disc is set to DefaultScalarPotential, and the PotentialType of each ring magnet is set to VectorPotential to support permanent magnet magnetization simulation.

[0010] Through the technical scheme, the potential type configuration of different components is clear, which avoids the failure of magnetization simulation caused by parameter setting error, is the key refinement of model attribute definition link, and ensures the accuracy of the simulation physical process.

[0011] As a further improvement of the above scheme, in step S51, the optimization range of the iron stand height is 0.7 to 1.5 times of the original height, and the S2-S4 process is repeated for each height value to determine the optimal iron stand height that minimizes ΔB through iterative simulation.

[0012] As a further improvement of the above scheme, in step S51, the iterative step of the iron stand height is set to 0.05 times of the original height, and the discrete values at least include 0.9H0, 1.0H0, 1.1H0, 1.15H0, 1.253H0, 1.3H0, wherein H0 is the original height of the iron stand.

[0013] Through the technical scheme, by limiting the step size and specific sampling points, the parameter scanning process of structure optimization is standardized, and the optimal height is avoided to be missed due to improper sampling interval.

[0014] As a further improvement of the above scheme, in step S52, the optimization material distribution ratio specifically includes: setting all sectors of the outermost circle and the second circle of the magnetic disc to NdFeB material, and implementing angular material replacement along the model predefined reference axis (such as X axis, 0° direction) symmetrically from the third to the sixth circle, that is, replacing 8 continuous sectors of the third, fourth and fifth circles to NdFeB, and replacing 4 continuous sectors of the sixth circle to NdFeB, and repeating the S2-S4 process to determine the optimal material distribution.

[0015] Through the technical scheme, by introducing a designed, non-uniform magnetic moment distribution, the magnetic field components generated thereby can cancel each other out in the target area with the inherent harmful harmonic components of the system, thereby achieving the effect of "targeted shimming".

[0016] As a further improvement of the above scheme, in step S52, the symmetric angle range of the angular material replacement is [-θ1, θ1] (third, fourth and fifth turns) and [-θ2, θ2] (sixth turn), wherein θ1 is determined by "the central angle corresponding to 8 continuous sectors", θ2 is determined by "the central angle corresponding to 4 continuous sectors", and the replacement azimuth synchronously covers the X-axis and Z-axis directions to realize bidirectional harmonic compensation.

[0017] Through the above technical scheme, steps S51 and S52 are sequentially executed, and the first stage provides an approximately optimal macroscopic magnetic environment for the second stage, and the second stage performs microscopic harmonic repair on this basis, and the two are coordinated to finally realize globally optimal magnetic field uniformity.

[0018] As a further improvement of the above scheme, in step S52, the target of the material distribution optimization is to offset the inherent quadrupole moment and sextupole moment low-order angular magnetic field harmonics of the disk system, and the ΔB of the target spherical region after optimization is reduced by at least 13% compared with before optimization.

[0019] Through the above technical scheme, the "harmonic suppression type" and the "effectiveness quantitative index" are clearly defined, so that the optimization target is measurable, and at the same time, the significant advantage that the ΔB of the target spherical region after optimization is reduced by at least 13% compared with before optimization is highlighted.

[0020] As a further improvement of the above scheme, the Br of the NdFeB material is ≥1.4T, the Br of the SmCo material is ≥1.1T, and in step S3, the MaximumElementSize of the grid division is set to 2mm for the magnet region and 5mm for the air region.

[0021] Through the above technical scheme, the material performance and the grid size are limited, which not only ensures the simulation authenticity of the magnetization intensity, but also balances the "simulation accuracy" and "calculation efficiency" through differentiated grid density, thereby improving the engineering practicability of the method.

[0022] Compared with the prior art, the beneficial effects of the present application are: 1. "Structural magnetic circuit regulation" and "material harmonic suppression" are clearly defined as two different levels of optimization problems that are related to each other, and a sequential solution scheme is proposed. In particular, the "angularly selective material replacement" strategy breaks the traditional symmetrical design mindset and directly starts from the physical principle of harmonic cancellation, which is strong in design.

[0023] 2. Through two-stage optimization, macroscopic and microscopic factors that affect magnetic field uniformity can be systematically solved. Practice shows that this method can significantly reduce the magnetic field uniformity deviation (ΔB) of the target spherical region by 13%, which is much better than single parameter optimization 3. The present application provides a complete and operable process guide from modeling, simulation to parameter optimization, which particularly elaborates the implementation steps on the professional platform of Opera3D, has strong engineering practicability and repeatability, and can directly guide engineering design.

[0024] 4. The core ideas of "first macrostructure shaping, then microsource distribution optimization" and "targeted compensation for harmonic components" contained in the present application not only apply to the specific disk model, but also have important reference significance and promotional value for the magnetic field uniformity optimization of other types of complex permanent magnet systems. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 is a three-dimensional model diagram of the disk system of the embodiment of the present application.

[0026] Figure 2 is a model diagram after finite element meshing in Opera3D.

[0027] Figure 3 is a magnetic field distribution cloud diagram of the central cross section of the disk system after optimization by the method of the present application (the color uniformity in the spherical region can be directly observed).

[0028] Figure 4 is the magnetic field intensity distribution curve along a certain diameter in the target spherical region before optimization.

[0029] Figure 5 is the magnetic field intensity distribution curve on the same path after optimization. DETAILED DESCRIPTION

[0030] In the following, the present application will be further described in combination with the drawings and specific embodiments, and it should be noted that the following described embodiments or technical features can be combined in any manner to form new embodiments without conflict.

[0031] Embodiment 1: Please combine Figure 1 - Figure 5 The implementation object of the present embodiment is a high-field double-disk system, and a single disk has a radius of 29.8 cm and is composed of seven rings of fan-shaped magnets with different heights (from outside to inside: 4.7 cm, 2.2 cm, 1.95 cm, 1.95 cm, 2.18 cm, 2.24 cm, and 2.06 cm), the height of the stand is determined to be 37.6 cm through pre-research, and the target optimization region is a spherical space with a radius of 10 cm at the geometric center of the two disks.

[0032] S1, accurate geometric modeling and data preparation Using SolidWorks 3D CAD software, an assembly model containing all components is created strictly according to dimensions. Key steps include: creating each ring of sector magnets individually (ensuring the correct sector angle and radius), eddy disk (thickness needs to be defined), and iron stand (its height will be used as a parametric variable). After completion, the model is exported in a neutral format such as STEP or IGES.

[0033] S2, Model Import and Physical Property Definition in Opera3D 1. Model Import and Entity Picking: Import the geometry file into the Modeller module of Opera3D, and use the PickBodies and PickEntity commands under the Model menu to precisely select the iron stand, vortex disk, and each ring of magnets as an independent volume.

[0034] 2. Material property assignment: Set physical properties for each volume.

[0035] Iron stand and vortex disk: Set its MaterialLabel to DT4 (or a custom label), PotentialType is usually set to DefaultScalarPotential, associate the predefined nonlinear BH curve data of DT4 material in VolumeDataLabel, and set an appropriate MaximumElementSize according to the size of the part.

[0036] Permanent magnets: Set a MaterialLabel for each magnet volume. PotentialType needs to be set to VectorPotential to support permanent magnet simulation. The key step is to define the magnetization direction: specify the magnetization vector for each magnet volume using SetVolumeProperties. It is also necessary to associate the correct permanent magnet material data (including Br, Hc).

[0037] S3. Solution Domain Setup and Mesh Generation 1. Define boundary conditions and free space: Use the ModelSymmetry command to set symmetric plane boundary conditions according to the symmetry of the model (such as 1 / 4 symmetry) to significantly reduce the amount of computation. Then, create a sufficiently large cylindrical free space to completely enclose the entire model to simulate an infinitely far boundary.

[0038] 2. Finite element mesh generation: First, execute CreateModelBody to generate the geometry for calculation, then click GenerateSurfaceMesh to generate the surface triangular mesh, check the mesh quality during this process to ensure that there are no abnormal cells in the key areas; finally, execute GenerateVolumeMesh to perform three-dimensional tetrahedral mesh division, and in the pop-up dialog box, set the Tolerance parameter to 1.0E-08 to ensure accurate discretization of geometric features, and for areas with large magnetic field gradients such as magnets and air gaps, ensure that the mesh is dense enough through local control.

[0039] S4. Solver settings and calculation In the Analysis type, select TOSCAMagnetic (static magnetic solver): create a solver database by CreateAnalysisDatabase, select Linear for the element type, and finally click the Prepare and Solve button to submit the calculation.

[0040] S5. Two-stage system optimization implementation First stage: iron stand height optimization 1. Take the iron stand height as a variable, and create new geometric models at its original height (denoted as H0) 0.9H0, 1.0H0, 1.1H0, 1.15H0, 1.253H0, 1.3H0, respectively. The geometric models can be reimported after parameterized modeling in CAD. In SolidWorks, modify the iron stand height to 27 cm (0.9H0), 30 cm (1.0H0), 33 cm (1.1H0), 34.5 cm (1.15H0), 37.6 cm (1.253H0), and 39 cm (1.3H0), respectively, and finally export 6 STEP models.

[0041] 2. For each height model, repeat the process of S2 to S4. After each solution, create dense field value sampling points within the target spherical domain in the Post-Processor, and calculate the peak-to-peak deviation ΔB of the magnetic field as shown in the table below:

[0042] 3. According to the above figure, plot the curve of ΔB with the change of iron stand height, and find that when the height is 1.253H0, ΔB is the smallest. Therefore, the optimal iron stand height H-opt is determined to be 1.253H0.

[0043] Second stage: material angular distribution optimization 1. Fix the model to the version with the iron stand height equal to H-opt.

[0044] 2. Implement the optimization strategy: Reference setting: based on the model of H-opt=37.6cm, modify the magnet material, set the material of all sectors of the 1st and 2nd rings to NdFeB, and simulate ΔB=0.0174.

[0045] Angular replacement: take the positive direction of the X-axis as the reference angle (0°), in the 3rd and 4th rings, select all sectors within the angle range [-θ1, θ1] (corresponding to 4 sectors on the left and right, the specific angle is calculated according to the number of sectors), change the material property from SmCo to NdFeB, in the 5th and 6th rings, select all sectors within the angle range [-θ2, θ2] (corresponding to 2 sectors on the left and right), and perform the same material replacement.

[0046] 3. Simulate the optimized model (S2-S4), and calculate ΔB, and compare it with full NdFeB, full SmCo, and only outer ring NdFeB inner ring SmCo symmetric distribution scheme, and calculate the optimized ΔB=0.0028, which is reduced by 13% compared with the initial ΔB=0.127, and reaches the optimization target.

[0047] 4. Effect verification: generate visual results by Opera3D to verify the optimization effect: Magnetic field cloud map: the magnetic field distribution cloud map of the central section after optimization ( Figure 3 ) shows that the color of the target spherical region is uniform, and there is no obvious magnetic flux gradient; Magnetic field curve: before optimization ( Figure 4 ) the Bz curve of the target region along the Y-axis fluctuates by 15T, and after optimization ( Figure 5 ) the fluctuation amplitude is reduced to 2T, and the fluctuation trend is gentle, which confirms that the angular harmonic is effectively suppressed.

[0048] The above embodiments are only preferred embodiments of the present application, and cannot be used to limit the scope of protection of the present application, and any non-essential changes and replacements made by those skilled in the art on the basis of the present application are within the scope of protection of the present application.

Claims

1. A method for improving the non-uniformity of magnetic field in space on a disk, characterized in that, Includes the following steps: S1. Geometric Modeling and Import: Construct an accurate model of the disk system, including seven fan-shaped magnets, parallel disks, vortex disks, and iron frame, based on 3D modeling software. Import the model into Opera3D finite element analysis software in STEP or IGES format. S2. Model Property Definition and Solution Settings: In Opera3D, define the material properties (the iron stand and eddy disk are associated with the DT4 material BH curve, and the magnets are associated with the remanent magnetization Br and coercivity Hc data of NdFeB / SmCo), magnetization direction (the magnets are set to axial magnetization), and boundary conditions (a 1 / 4 symmetry plane boundary is set based on the model symmetry) for the iron stand, eddy disk, and each ring of magnets. Select the TOSCAMagnetic module to solve the static magnetic field. S3. Mesh Generation and Solution: Create a cylindrical air domain that encloses the model, generate surface triangular meshes and three-dimensional tetrahedral meshes, and perform solution calculations to obtain the system's magnetic field distribution. S4. Post-processing analysis: Create dense sampling points in the target spherical region at the geometric center of the two disks, extract the magnetic induction intensity Bz component, and calculate the magnetic field non-uniformity deviation ΔB=(B_max-B_min) / B average as the non-uniformity index. S5, Two-stage system optimization: S51. Structural parameter optimization stage: With the goal of reducing the magnetic field inhomogeneity in the target area, the height parameter of the iron frame is optimized, and its optimal value is determined through iterative simulation. S52. Material Distribution Optimization Stage: Based on the optimal iron frame height, optimize the angular and radial distribution ratio of NdFeB and SmCo materials in the seven-ring disk magnet, and determine the optimal distribution through iterative simulation.

2. The method for improving the non-uniformity of magnetic field in a disk space according to claim 1, characterized in that, In step S1, the 3D modeling software is SolidWorks. When modeling, the radius, sector angle and height of each sector magnet need to be defined separately, and the initial height of the iron frame model is 30cm.

3. The method for improving the non-uniformity of magnetic field in a disk space according to claim 1, characterized in that, In step S2, the PotentialType of the iron frame and the eddy current disk is set to DefaultScalarPotential, and the PotentialType of each ring of magnets is set to VectorPotential to support permanent magnet magnetization simulation.

4. The method for improving the non-uniformity of magnetic field in a disk space according to claim 1, characterized in that, In step S51, the optimized range of the height of the iron frame is 0.7 to 1.5 times its original height. The process of S2-S4 is repeated for each height value, and the optimal height of the iron frame that minimizes ΔB is determined through iterative simulation.

5. The method for improving the non-uniformity of magnetic field in a disk space according to claim 4, characterized in that, In step S51, the iteration step size of the height of the iron frame is set to 0.05 times the original height, and the discrete values ​​include at least 0.9H0, 1.0H0, 1.1H0, 1.15H0, 1.2H0, and 1.3H0, where H0 is the original height of the iron frame.

6. The method for improving the non-uniformity of magnetic field in a disk space according to claim 1, characterized in that, In step S52, the optimization of the material distribution ratio specifically includes: setting all sectors of the outermost and second rings of the disk magnet to NdFeB material, performing angular material replacement symmetrically along the predefined reference axis of the model for the third to sixth rings, that is, replacing 8 consecutive sectors with NdFeB in the third, fourth and fifth rings, and replacing 4 consecutive sectors with NdFeB in the sixth ring, and repeating the S2-S4 process to determine the optimal material distribution.

7. The method for improving the non-uniformity of magnetic field in a disk space according to claim 6, characterized in that, In step S52, the symmetrical angle range of the angular material replacement is [-θ1, θ1] (third, fourth, and fifth circles) and [-θ2, θ2] (sixth circle), where θ1 is determined by "the central angles corresponding to 8 consecutive sectors" and θ2 is determined by "the central angles corresponding to 4 consecutive sectors". The replacement orientation synchronously covers the X-axis and Z-axis directions to achieve bidirectional harmonic compensation.

8. The method for improving the non-uniformity of magnetic field in a disk space according to claim 1, characterized in that, In step S52, the goal of optimizing the material distribution is to counteract the inherent quadrupole and hexapole low-order angular magnetic field harmonics of the disk system, and the ΔB of the target spherical region after optimization is reduced by at least 13% compared with that before optimization.

9. A method for improving the non-uniformity of magnetic field in space on a disk according to any one of claims 1-8, characterized in that, The Br≥ of the NdFeB material 1.4T, Br≥1.1T for SmCo material, and the MaximumElementSize of the mesh in step S3 is set to 2mm for the magnet region and 5mm for the air region.

Images