WO2023140753A1

WO2023140753A1 - Method for manufacturing segmented permanent magnets from low-grade magnetically hard sintered raw material

Info

Publication number: WO2023140753A1
Application number: PCT/RU2023/000011
Authority: WO
Inventors: Александр Метталинович ТИШИН
Original assignee: Общество С Ограниченной Ответственностью "Ампермагнит"
Priority date: 2022-01-19
Filing date: 2023-01-17
Publication date: 2023-07-27

Abstract

A method for treating low-grade magnetically hard sintered raw material for the manufacture of segmented permanent magnets (SPMs) includes carrying out primary sorting of specimens of raw material into groups according to geometric characteristics and external appearance, taking into account desired SPM parameters; obtaining sample measurements of magnetic, chemical and microstructural parameters and performing a HAST test to determine the possibility in principle of manufacturing an SPM; and manufacturing a pilot batch of SPMs, which includes cutting the specimens into parts having a set configuration, gluing the parts into a multi-layer stack, cutting and grinding the surface of the obtained multi-layer stack to a set size, and applying a protective coating to the surface of the obtained SPM. If the SPM does not meet the desired characteristics, the specimens of raw material are subjected to secondary sorting into sub-groups according to magnetic, chemical and microstructural parameters and, during manufacture, the arrangement and orientation of individual parts with different magnetic characteristics within the multi-layer stack under formation are additionally selected with a view to obtaining the desired overall magnetic characteristics. The result is the efficient treatment of recyclable material and a broadening of the range of semi-finished products.

Description

METHOD FOR MANUFACTURING SEGMENTED PERMANENT MAGNETS FROM OFF-CONDITIONAL MAGNETIC HARD SINTERED RAW

Technical field

The invention relates to the processing of secondary raw materials and can be used in the manufacture of permanent magnets with specified parameters.

Prior Art

It is known that in the process of manufacturing, transporting and installing rare-earth magnets (REM) manufactured by powder metallurgy, a large number of defective magnetic parameters, dimensions, chips, appearance and protective coating of various magnetic elements are formed. These also include magnetically suitable, but having a different size, as well as previously used magnetic elements. Rare earth resources are non-renewable resources. In order to save resources, reduce industrial waste and protect the environment, these raw materials and waste must be reused, and this can even bring significant social and economic benefits.

Methods for REM recycling by means of various chemical-metallurgical processing methods are described (see the review by Yongxiang Yang, et al., REE Recovery from End-of-Life NdFeB Permanent Magnet Scrap: A Critical Review, 2016, J. Sustain. Metall (DOI 10.1007/s40831-016-0090-4). So, in the invention (RU 209733 0 Cl, Karmannikov, 27.1 1.1997) describes the isolation of compounds of rare earth elements from production waste and the use of permanent magnets. It includes heat treatment of waste at 80-700 ° C and their dissolution in nitric acid. From the resulting solution containing 0.4-3.0 M nitric acid, the amount of REE is extracted into the organic phase. This organic phase is used as the initial solution for the isolation of neodymium from it by extraction of other REE E with TBP The resulting solution of neodymium nitrate contains less than 0.05 wt.% other REE. Extraction of neodymium from waste is at least 95%. In the invention JP 2020084331 (A), INST JOZEF STEFAN, 06/04/2020, a more environmentally friendly and economical method for recycling scrap NdFeB magnets at the end of their service life by anodizing is described. Other chemical methods for the reuse of sintered galvanized neodymium, iron and boron wastes have also been described (see, for example, CN 108188146 (A), JINGCI MAT SCIENCE CO LTD, 06/22/2018). Along the way, technological methods are considered for detaching used REM from equipment elements by heating in organic solvents (JP 2012064889 (A), KOBE STEEL LTD, 03/29/2012), however, the magnets themselves are not modified and are used "as is".

At the same time, such methods are energy-intensive, harmful to the environment, and in addition, they contribute to the complete loss of magnetic properties, and magnets with benign characteristics fall into the redistribution. It should be especially noted that the metallurgical processing completely destroys the microstructure and stored magnetic energy, the creation of which has already consumed energy resources for heat treatment and magnetization.

Known manufacturing technology of a composite multilayer NdFeB magnet (see, for example, US 9782953 (B2), YANTAI SHOUGANG MAGNETIC MAT INC., 10/10/2017; CN 201745223 U, YANTAI ZHENGHAI MAGNETIC MATERIAL CO Ltd., 02/16/2011). The technology involves selecting and cleaning a plurality of permanent magnets, phosphating, rinsing with alcohol to remove grease and rust after phosphating, plasma cleaning process, applying a layer of insulating adhesive with a thickness of 1 µm to 100 µm on the surface of each of the NdFeB permanent magnets, applying a predetermined clamping pressure in the range of 0.1 MPa to 10 MPa, curing, processing to a predetermined size, applying a protective layer of epoxy and phenolic resin with a thickness of 8 microns to 50 microns. However, the magnetic characteristics of both composite elements and composite magnets obtained from them are not considered. At the same time, the selection of magnetic parameters, the shape and size of each of the layers of the composite magnet, as well as the use of not only NdFeB, but also SmCo REM at the same time, makes it possible to provide a given distribution magnetic field in the required volume and increased temperature nature of the truth.

Known methods of processing substandard permanent magnets neodymium-iron-boron (NdFeB) into elements with specified magnetic and geometric characteristics, but without metallurgical processing. In the event that the magnetic characteristics have low operational parameters, the magnets are subjected to local surface diffusion of dysprosium and/or terbium, followed by heating. The analysis of such procedures is described in particular in the review of the prior art in patent RU 2423204 C2, INTERMETALLIX CO. LTD. (JP), 07/10/2011.

Closest to patentable is the method described in the invention JP 2002141239 (A), H1ROSE YOICH1, 05/17/2002 - prototype. The method of regenerating magnets consists in marking and then cutting the substandard magnet into parts, and then gluing these parts in a certain order into a new powerful magnet. At the same time, it is noted that, compared with the metallurgical method, the energy required to cut the magnet is small, and routine technology can be used, and the material loss consists only in the cutting allowance. Accordingly, this technology makes it possible to obtain a new powerful REM with a high yield without consuming a large amount of energy.

However, the high consumer qualities of such a regenerated magnet are primarily determined by the quality of the raw materials - the initial parameters of the microstructure of the hard magnetic material, weight loss (relative to the initial one), or corrosion, which can lead to an irreversible change in magnetic properties. This part requires a detailed study of representative samples of the raw materials used, which requires large time and human resources and unique scientific equipment (for example, an analytical high-resolution electron microscope with an objective aberration corrector and EDX and EELS attachments), not intended for serial factory research. This is due to the fact that the magnetic properties are determined by a number of microstructure parameters, and not only by the properties of the grains themselves, but also by a number of magnetic, electronic, structural and chemical properties of their boundaries (phases enriched with neodymium), etc. characteristics, and primarily with the magnitude of the coercive force, requires further detailed study (TG Woodcock, W. Zhang, G. Hrkac, G. Ciuta, NM Dempsey, T. Schrefl, O. Gutfleisch and D. Givord. Understanding the microstructure and coercivity of high performance NdFeB-based magnets - Scripta Materialia 67 (2012), 536-541). Therefore, the identification of really suitable magnets is extremely difficult.

In addition, the prototype is limited to the rectangular shape of regenerated magnets with the same magnetic properties (of the same brand). At the same time, it is known that one of the ways to increase the magnitude and concentration of the magnetic induction created by a system of permanent magnets (reduce the magnitude of the field outside the magnetic system) is, for example, the use of Halbach cylinders (see, for example, RU 2466491 C2, LLC "Perspective magnetic technologies and consultations", 10.11.2012). However, the wide use of such systems is limited by the fact that the magnetization vector of the cylinder material should ideally rotate smoothly in the cylinder of the magnetic system being created. In practice, such systems are segmented and consist of parts in which the magnetization vector rotates abruptly when moving from one part to another. But even the manufacture of such systems, as a rule, is carried out by cutting in the right direction (the direction of magnetization) of each such part from identically magnetized parallelepipeds, which leads to a large number of trimmings and significantly increases the cost of manufacturing permanent magnets (from 30 to 100% or more) for the end user.

Disclosure of invention

The present invention is aimed at solving the problem of increasing the efficiency, completeness of processing and expanding the range of secondary raw materials - substandard elements of permanent magnets: defective in magnetic parameters, dimensions, chips, appearance and protective coating. Such raw materials include those suitable in terms of magnetic characteristics and microstructure, but having an inappropriate size, as well as previously used permanent magnets and systems based on them. This is of particular importance for mass production (with a volume of 100-500 tons per year) of the same large (weighing 0.1-1.0 kg or more) permanent magnet with the same geometry and magnetic properties. Rejection due to chips or defects in the coating can occur at the final stages of production (grinding and coating) of a practically finished sintered magnet blank or decommissioned devices (for example, wind turbines) using 0.6-10 tons of such permanent magnets. The above sharply reduces the efficiency of production (for example, the yield of finished products).

The patented method of processing hard magnetic sintered raw materials for the manufacture of segmented permanent magnets (SPM) consists in marking and then cutting into parts of a substandard magnet, and then gluing these parts in a certain order into a new powerful magnet.

Primary sorting of samples of raw materials into groups according to geometric characteristics, storage conditions and appearance is carried out taking into account the required parameters of the SPM and the allocation of raw materials for subsequent metallurgical processing. Demagnetization is carried out if necessary, selective measurement of magnetic, chemical and microstructural parameters, a HAST test and comparison with the initial drawing (if any) for each of the sorted groups of raw materials with the determination of the fundamental possibility of manufacturing SPM.

Selection of the first (basic) technology and production of an experimental batch of SPM from each sorted group suitable for processing, including demagnetization if necessary, cleaning the surface of samples, cutting samples into parts of a given configuration, gluing parts into a multilayer package, cutting and grinding the surface of the obtained multilayer package to a given size and applying a protective coating to the surface of the obtained SPM. The integrated magnetic characteristics of the obtained SPM, as well as systems based on them, are measured and compared with the required characteristics.

If SPM and systems based on them correspond to the required characteristics, secondary sorting of raw material samples is carried out into subgroups according to magnetic, chemical, microstructural parameters and HAST test for the selected first technology, including demagnetization if necessary, cleaning the surface of samples, cutting samples into parts of a given configuration, gluing parts into a multilayer package, cutting and grinding the surface the resulting multilayer package to fit the size and applying a protective coating to the surface of the obtained SPM.

If the SIM does not meet the required characteristics, the secondary sorting of the raw material samples is carried out into subgroups according to magnetic, chemical, microstructural: parameters and the HAST test for the selected second technology, including demagnetization, if necessary, cleaning the surface of the samples, cutting the samples into parts of a given configuration, selecting the layout and orientation of single parts with different magnetic characteristics in the formed multilayer package from the condition of obtaining the required integral magnetic characteristics, gluing the parts into a multilayer package, cutting and polishing ovulation of the surface of the resulting multilayer package to size and applying a protective coating to the surface of the obtained SPM.

The method can be characterized by the fact that the neodymium-iron-boron and samarium-cobalt raw materials include magnetic elements defective in terms of magnetic parameters, dimensions, chips, appearance and protective coating, as well as suitable magnetic characteristics, but having a different size, as well as previously used permanent magnets.

The method can also be characterized by the fact that the raw material for subsequent metallurgical processing, including hydrometallurgy methods, includes waste from cutting and grinding parts and a glued multilayer package, as well as broken permanent magnets.

The method can be characterized, in addition, by the fact that the magnetic parameters include, but are not limited to, the parameters of the demagnetization curve at temperatures T=+20 and +150C: coercive force in induction Hcb and magnetization Hcj, residual induction Br, maximum energy product (BH)max, operating temperature range, Curie temperature Tc, as well as temperature coefficients a of reversible changes in the values aBr, aHcb, aH cj, as well as the direction of the axis of the magnetic texture, flux linkage, magnetic dipole moment, the nature of the distribution of the magnetic field induction created in the simulator of the magnetic system of the final product.

The method can also be characterized by the fact that single parts with different residual induction characteristics are made from raw materials different grades, while the number of grades of raw materials included in one SPM is from 1 to 5.

The method can also be characterized by the fact that after cutting the samples, the parts are subjected to the procedure of local surface diffusion of dysprosium and/or terbium.

The method can also be characterized by the fact that cutting samples into parts of a given configuration, as well as the surface of a multilayer package to size, is carried out by an electric spark or hydroabrasive method and / or a diamond disk, followed by mechanical and / or chemical elimination of the products of the specified machining.

The method can also be characterized by the fact that cutting is carried out sequentially in two or more stages from the condition of minimizing chips in the sawing zone and the required direction of magnetization, while at the first stage cutting of samples connected to each other is carried out, and at the second stage - cutting of the obtained parts of a given size for gluing them into a multilayer package, and at the third stage - cutting to form the required direction of the magnetic texture axis for magnetization of the SPM.

The method can also be characterized by the fact that the specified configuration of parts is selected from the list, including, but not exhaustive: a plate, a parallelepiped, a trapezoid plate, a cylinder, a part of a hollow cylinder, including for Halbach cylinders.

The method can also be characterized by the fact that parts are glued into a multilayer package with heat-resistant glue, while the glue is applied to both glued surfaces twice with intermediate drying, followed by joining the parts into a package and keeping the package under load until the glue cures, and also by the fact that a protective coating with a hardness of H6 on the surface of the SPM multilayer package is made of an epoxy compound with a thickness of 10 to 35 microns.

EFFECT: increased efficiency, completeness of processing of secondary magnetic raw materials and expansion of the range of semi-finished products, including defective ones in terms of magnetic parameters, dimensions, chips, appearance and protective coating of elements of various magnetic systems, as well as previously used permanent magnets for the manufacture of effective segmented permanent magnets and magnetic systems with specified magnetic and geometric characteristics.

The patented processing method makes it possible to cut parts from rare-earth metals of any shape, including: plates, trapezoidal elements and other shapes, for example, for Halbach cylinders, and thus ensure a significant reduction in the cost of manufacturing the final product. Integratively, this will lead to a significant reduction in weight and size characteristics (due to the concentration of the magnetic field in the working gap), as well as an increase in efficiency (reduction of losses due to Foucault currents) of products based on permanent magnets. When choosing the thickness of the parts to be cut out (for example, plates), it was taken into account that the inhomogeneities in the distribution of magnetic induction that occur in the magnetic system due to gluing parts from different raw materials into a multilayer SPM package will be smoothed out by a magnetic circuit (for example, a rotor) and should not lead to an increase in the magnitude of the torque ripple in electric drives, additional voltage ripples and gramonics in electric generators and other additional effects affecting the normal operation of electric machines.

In addition, as a rule, when reusing the SPM, questions arise as to whether the SPM was subjected to mechanical (vibration), electromagnetic, temperature, and other environmental influences during operation. The result of such impacts may be a violation of the microstructure of the hard magnetic material, weight loss (relative to the initial one) or corrosion, which can lead to an irreversible change in the magnetic properties and the impossibility of reprocessing by the claimed method. If the presence of weight loss and traces of corrosion can be determined by well-known methods, then, as noted above, to determine subtle changes in the microstructure, careful, long-term and labor-intensive studies are required on a large volume of samples using unique equipment.

At the same time, since the microstructure is directly related to the magnitude of the REM coercive force and its temperature dependence, the following simplified method for determining the suitability of the specified magnetic raw material for further processing is proposed:

- external inspection of raw materials (conditions of storage and/or operation, absence of visible traces of corrosion and/or damage to the coating);

- selection of undamaged magnets with undisturbed geometric dimensions, appearance and preserved coating of at least one protective layer;

-control studies of the microstructure, chemical and phase composition of raw materials on one or three samples of raw materials, depending on the batch size. Comparison with witness samples or manufacturer's output control data (if available). A typical microstructure for sintered Nd-Fe-B magnets: Nd2Fel4B grains about 5 µm in size are surrounded by thin layers of oxide and metal phases. enriched with Nd with a thickness of 0.5-3 nm in a volume of less than 10%.

If deviations from the required parameters are found, an increased double sampling of samples and repeated studies are recommended;

- random sample for HAST tests in the amount of not more than 0.1% of the batch. If deviations from the required parameters are found, an increased double sampling of samples and repeated studies are recommended.

- random sampling for measurements of the coercive force and its dependence on temperature in a volume of not more than 0.1% of the batch. If deviations from the required parameters are found, an increased double sampling of samples and repeated studies are recommended.

Subject to passing these tests, consider the raw material suitable for processing. The processing method is determined by the remaining magnetic characteristics of the raw material and the requirement for the final product.

Brief description of the figures of the drawings Fig.1 - flowchart of the method; figure 2 - orientation cutting of samples of raw materials on the details; fig.Z - multilayer package when gluing parts of raw materials of the same brand; figure 4 - multilayer package when gluing parts of raw materials of two different grades; figure 5 - multilayer package when gluing parts of raw materials of three different grades; Fig.6 - characteristics of the magnet NdFeB brand N4211 at 20C; fig. 7 - the same as in Fig.6, for brand N5 OM; fig. 8 - characteristics of SPM from alternating plates of grades N42H and N50M at 20C.

Embodiments of the invention

The process of processing substandard magnetic raw materials begins with numerical simulation of the magnetic parameters and dimensions of the SPM, which will be used, for example, to manufacture the final product.

Further, primary sorting of samples of raw materials into groups according to geometric characteristics and appearance is carried out, taking into account the required parameters of SPM. A selective measurement of magnetic, chemical and microstructural parameters is carried out and a HAST test is carried out for each of the sorted groups of raw materials with the determination of the fundamental possibility of manufacturing SPM.

Table 1 shows the requirements for the SIM based on the results of numerical simulation and the results of testing the parameters of magnets from different manufacturers. The magnetic parameters include, but are not limited to, the parameters of the demagnetization curve at temperatures T=+20 and +150C: coercive force by induction Hcb and magnetization Hcj, residual induction Br, maximum energy product (BH)max, operating temperature range, Curie temperature Tc, as well as temperature coefficients a of reversible changes in the values aBr, aHcb, aHcj (GOST R 52956-2 008 "HARD MAGNETIC SINTERED MATERIALS BASED ON NEODIMIUM-IRON-BORON ALLOY"). Some of these parameters are shown in Table 1.

Selective measurement of magnetic, chemical, microstructural parameters and HAST test is carried out in a volume of not more than 0.1% of the number of samples of sorted groups of raw materials, and if deviations from the required parameters are detected, samples are taken in a volume of 0.2% and repeated studies. Average measured results by samples are presented. Table 1

Requirements for SPM based on the results of numerical simulation and the results of testing the parameters of magnets from different manufacturers

Next, an experimental batch of SPM is manufactured from each sorted group according to the first technology.

The surface of the samples is cleaned, the samples are cut into parts of a given configuration, and the parts are glued into a multilayer package.

In FIG. 2 shows the orientation cutting of samples 1 of raw materials into parts 2, the direction of the axis 3 of the magnetic texture. Cutting specimens on a predetermined part configurations are carried out by an electrospark or hydroabrasive method or a diamond disk. On fig.3 shows the structure of a multilayer package obtained by gluing compound 4 parts 20 raw materials of the same brand. Pos. 5 shows the surfaces of the package that are subjected to subsequent processing to size.

In FIG. 4 shows the structure of a multilayer package obtained by gluing parts 20 and 21 of two different brands.

In FIG. 5 shows the structure of a multilayer package obtained by gluing parts 20,21,22 of raw materials of three different grades.

Then the surface of the obtained multilayer package is cut and polished to a predetermined size and a protective coating is applied to the surface of the obtained SIM. A detailed description of the technology is given below.

1. In accordance with the results of numerical simulation of the digital electromagnetic twin of the final product, manufactured SPMs must have the magnetic characteristics presented in the top line of Table 1 "Requirements for the parameters of the SPM based on the results of numerical simulation", and also meet the following geometric requirements:

1.1. finishing geometric dimensions after gluing, re-cutting and final grinding 2.60 ^x 21.33 ^x 30.00 mm (tolerance for all dimensions without taking into account the coating thickness ± 0.1 mm), t . the rotor with an axial length of 150 mm will consist of five identical SPMs of 30 mm each;

1.2. thickness and orientation of the plates included in the final SPM after re-cutting 2.60 ± 0.10 x 21.33 ± 0.10 ^x 5.00 + 0.007-0.05 mm;

1.3. The thickness and dimensions of the plates glued before the first cutting: 21.33 ± 0.10 ^x 5.00 + 0.00 / -0.05 x 86.50 ± 0.10 (cut from the original raw material). In this case, the thickness and size of the plates for gluing is selected taking into account the maximum utilization of the geometric dimensions of the available raw materials.

Samples of raw materials were sorted into groups according to geometric characteristics and appearance, taking into account the required parameters of SPM. Separately allocated waste magnetic raw materials for subsequent metallurgical processing.

As a result of sorting, two groups of raw materials were identified: group 1 - about 100,000 pieces of used permanent magnets; group 2 - about 24,000 pieces of permanent magnets rejected at the input control due to the presence of chips that exceed the allowable dimensions.

All magnets have magnetic characteristics close to the required ones, produced by different manufacturers from different raw materials in different years, with slightly different chemical composition.

Group 1 source of raw materials - decommissioned wind turbine rotor poles, which initially had a five-layer protective coating: Ni/Cu/Ni, Sn and epoxy compound. The magnets are visually inspected. The appearance, the presence of chips, the safety of the coating, traces of corrosion as a result of storage, etc. are determined and fixed. It has been established that the raw material has a different degree of preservation and damage to the coating that occurred during the extraction of permanent magnets from the rotors of wind turbines. The above suggests that the feedstock was subjected to heat during extraction from the rotors, which could lead to delamination and damage to the Ni/Cu/Ni, Sn coating, demagnetization when heated above the Curie temperature (more than 312C), and a change in the microstructure of the feedstock when heated under operating conditions. Part of the raw material was subjected to mechanical stress.

Inspection of raw materials of group 2 established that all raw materials have a solid, unbroken coating (including chips), have no signs of corrosion, mechanical or thermal effects, and are suitable in appearance for further processing by the claimed method.

2. Primary sorting of permanent magnets of group 1 into the following subgroups is carried out: subgroup 1.1 - heavily damaged permanent magnets (broken); subgroup 1.2 - with significant chips, defects and traces of corrosion, however, allowing cutting samples into parts of a given configuration, outside the areas of chips and traces of corrosion; subgroup 1.3 - intact magnets with undisturbed geometric dimensions, appearance and preserved coating of at least one protective layer of Ni, Cu, Ni or Sn (without traces of corrosion).

Subgroup 1.1 is sent for processing by hydrometallurgy. Subgroup 1.2 is sent for additional sorting, taking into account the feasibility of processing using the proposed method. After additional sorting, 40% of raw materials of subgroup 1.2 were transferred to subgroup 1.1. The remaining 60% were transferred to subgroup 1.3. After that, a random sample of magnets is made from groups 1.3 and 2 to study magnetic and geometric parameters, microstructure and other parameters. A random sample is made in the amount of not more than 0.1% of the lot. The initial information on the magnets used in the manufacture of the wind turbine is taken into account, as well as the results of previous tests.

3. A study of the properties of permanent magnets is carried out in laboratories that have specialized accreditation and the corresponding verified equipment entered in the state register. The list of equipment is presented in Table 2.

Table 2.

List of equipment for basic measurements and testing of magnets

To obtain additional information on the microstructure, chemical and phase compositions, selective studies can be carried out using the following equipment: scanning electron microscope (Carl Zeiss CrossBeam 1540EsB), multifunctional scanning probe microscope (FemtoScan), transmission electron microscope (JEOL JEM-2010), diffractometers DRON-4-07, RIGAKU Ultima IV, high-resolution electron microscope Titan 80- 300 and other means and methods. The microstructure (crystallite size, phase composition, as well as the structure, magnetic and electronic properties of neodymium-enriched interlayers at grain boundaries, etc.) determines the formation of a high-coercivity state. Determination of the elemental composition of the surface layers can be carried out using the EPMA-SEM Camebax analyzer (CAMECA, Gennevilliers, France). Additional control of the chemical composition - on the X-ray fluorescence spectrometer PRIMUS II.

All batches of raw materials pass the HAST-test according to the 1EC.60068-2-66:1994 standard on 10x10mm discs, uncoated, 240 hours at a humidity of 95%, a temperature of 132C and a pressure of 2.7 atm. The surface of test specimens may have traces of corrosion, material residues from electric spark cutting, therefore, before testing, preparation is carried out as follows:

1) the entire surface is polished until the products of electric spark cutting are removed;

2) cleaning is carried out in an ultrasonic bath in an ethanol medium;

3) wiped with filter paper;

4) the discs are dried with a hair dryer. Two samples from subgroups 1.3 did not pass the HAST test (mass loss was 55 and 38 mg/cm ² ). All samples from Group 2 passed the test. The repeated enhanced sampling confirmed the initial results. The batches of raw materials that did not pass the HAST test from subgroup 1.3 were deemed unsuitable for processing according to the claimed method and sent to group 1.1.

On the remaining batches, the analysis of the results obtained is carried out for the compliance of the obtained properties with the required magnetic and other parameters obtained from the results of numerical simulation. The degree of change in the microstructure, chemical and phase composition is established. Particular attention is paid to measurements of the coercive force, which is directly related to the state of the microstructure of the sample. The suitability of raw materials for processing is determined by their chemical and phase composition and the required microstructure. The obtained values of the magnetic parameters are presented in Table 1.

Due to the large spread of magnetic characteristics, an increased selection of magnets is made in double size. Repeated measurements of the magnetic characteristics are carried out, which confirm that REM manufacturers W, 1 and T have a residual induction at the lower limit of permissible values. Raw materials of manufacturers I and T have a value of Hcj in the range of 1324-1331 kA/m at a temperature of 20C, which is less than the value required for the normal operation of the electric generator 1355 kA/m. Permanent magnets of these manufacturers have significantly underestimated values of (VN)max in the range from 312 to 318 kJ/m ³ and overestimated values of аВг - in the range from -0.130 to -0.1 1%/°С.

One of the possible reasons for the spread of magnetic characteristics may be a violation of the microstructure of raw materials due to operation at high temperatures or the complex effect of environmental factors. Batches of raw materials from these manufacturers are sent for additional microstructure studies. Studies of the microstructure did not reveal significant deviations; therefore, it was concluded that the underestimated values of the coercive force are the result of a manufacturing defect at the stage of production of raw materials. The conclusion is made about the suitability of this raw material for further processing by the claimed method. In accordance with GOST R 52956-2008, the following values of temperature coefficients of the characteristics of NdFeB REMs are allowed: by induction Bg: ABg x 100/(Br x AT) from -0.12 to -0.8%/K; magnetization coercive force Hcj: AHcj x 100/(Hcj x AT) from -0.59 to -0.45%/K. These parameters greatly affect the performance of electromagnetic devices based on REM permanent magnets at high temperatures. To understand the nature of the influence of the above parameters on the operation of the final device, not individual characteristics of the raw material are considered, but their totality as a whole, and additional numerical simulation is carried out. That is, numerical modeling takes into account that if a particular permanent magnet according to the drawing at 20C has a value of Bg = 1.32T and a temperature coefficient by induction of 0.11% / C, then at a winding temperature of, for example, 150C, the value is Bg (150) = 1.131T.

But, if we take such a combination of parameters of the same raw material, for example, at 20C the value of Br is 1.28 T and the temperature coefficient for induction, which goes beyond the scope of GOST R 52956-2008 - 0.14% / C, then at a winding temperature of 150C the value Br (l 50) = 1.049 T. That is, the difference in Bg for these two magnets at a temperature of T = 150C will be about 10%.

The numerical simulation also takes into account the detected deviation in the value of the temperature coefficient for the coercive force, the value of which at the operating temperature directly affects the energy that the magnet gives to the external space. It has been established that the simultaneous action of these magnetic characteristics gives an integral effect and has a significant decrease in the induction in the air gap to 4-7% of the value for manufacturers W, I and T at an operating temperature of 200C.

Thus, according to the simulation results, subgroup 1.3 and 2 was divided into two new subgroups A and B. Subgroup A consists of raw materials 1.3 (raw materials produced Y-K-Z) + 2 and subgroup B consists of raw materials 1.3 (raw materials produced-W-I-T). Accordingly, individual technologies for further processing were selected for each subgroup.

The creation of a digital electromagnetic twin of a specific electric machine allows you to determine the exact nature of the influence of the identified magnetic deviations on the performance characteristics of the final product and determine the most effective measures to utilize the maximum amount of raw materials, including, if necessary, to take measures to adjust the coercive force and temperature coefficients.

The patented method makes it possible to manufacture an SPM that ensures operability under conditions of repeated overheating up to temperatures near the ZEP. For this purpose, single parts glued into a multilayer SPM package are made of NdFeB and SmCo grades, while a higher-temperature Sm2Col7 magnet can keep less temperature-resistant NdFeB from demagnetization in a wider temperature range. So, the NdFeB magnet of the N42AFI brand, manufactured by Yantai Shougang magnetic materials, Inc, has the following parameters at T=20C: Br=1.317T, Hcb=1032 kA/m, Hcj=2900 kA/m, (BH)max=342 kJ/m ³ . At the same time, serially produced magnets of the Sm2Col7 brand have lower parameters at T=20C: Bg=1.106 T, Hcb=833 kA/m, Hcj=2101 kA/m, (BH)max=233 kJ/m ³ .

Accordingly, after gluing two plates with a thickness of 0.5 mm, made from the indicated brands of magnets, the SPM will have a lower Br value of about 1.21 T. However, at higher temperatures, the characteristics are leveled off: NdFeB magnet N42AH at T=180C has the following parameters: Br=1.039T, Hcb=790 kA/m, Hcj=1018 kA/m, (BH)max=206 kJ/m ³ and can only be operated up to a temperature of 230C.

At the same time, serially produced magnets of the brand SmCo (SmiCop) can be operated up to temperatures of 350C and have very similar parameters at T=180C: Bg=1.035T, Hcb=769 kA/m, Hcj=l 518 kA/m, BHmax=201 kJ/m'. At the same time, it should be taken into account that a slight overheating of the copper wire of the stator of the electric machine above the insulation service temperature (for example, up to 250C with an insulation class of 200C) does not lead to an irreversible failure of the electric machine, but only logarithmically reduces the insulation service life). At the same time, even a single heating of the N42AH magnet above a temperature of 230C can lead to irreversible partial or complete demagnetization of the SIM. Therefore, for responsible applications, it is extremely important to exclude the possible complete or partial demagnetization of the SPM in the composition of the product during a single heating above 230C. Best Modes for Carrying Out the Invention

The processing technology is given below using the example of NdFeB and SmCo magnetic raw materials, however, the described technology can also be applied to magnetic raw materials, as well as to other magnetic illiquid products of both powder and classical metallurgy, for example, UNDK brand magnets. In the latter case, depending on the magnitude of the demagnetization fields in the magnetic system, it is possible to perform, for example, inserts from UNDK into a sintered NdFeB magnet.

Example 1. Raw materials of subgroup A were selected based on the results of research using the methods described above. The samples passed the HAST tests and meet the requirements for raw materials with the required magnetic properties. It was decided to rework this subgroup under the First Technology according to the following program:

- selection of an experimental batch weighing 5 kg;

- carrying out cleaning of epoxy coating residues in organic solvents (due to the high chemical activity of REE, it is carried out immediately before the start of the processing procedure);

- mechanical removal of the Ni/Cu/Ni/Sn coating by grinding;

- removal of grinding marks and transfer of raw materials to an inert environment to avoid corrosion;

- application of a transport coating by surface passivation: washing with water, then treatment with a solution of nitric acid (30-40C), then drying in an air stream of 100-120C);

- cutting on an electroerosive wire-cutting machine of a jet type (model DK7740-ME12) of plates to a size of 21.33 ± 0.10 x 5.00 +0.00 / -0.05 x 86.50 ± 0.10 (corresponding to paragraph 1.3);

- processing and cleaning the surface of the obtained plates before gluing in the following sequence: grinding the surface before removing the products of electroerosive cutting and control measurement of dimensions with a micrometer; cleaning in an ultrasonic bath in ethanol; wiping the surface of the obtained parts with filter paper and drying with a hairdryer with carrying out operations in calico gloves; - gluing the obtained parts into a multilayer package using heat-resistant adhesive VS-YuT (GOST 22345-77) in the following sequence: the adhesive is diluted with ethyl alcohol in a ratio of 1:3 and sprayed on both surfaces to be glued. Incubated for 1 hour at room temperature, then placed in an oven heated to (180 ± 5) C and maintained at this temperature for an hour. After that, a second layer of glue is applied with a spray gun, which is again dried for another 1 hour. After that, the plates are connected and placed in a clamp with a bonding force of the order of 2.5 kgf/cm ² . The press with glued elements is placed in a heating cabinet heated to a temperature of (180 ± 5) C and maintained at this temperature for 2 hours. To increase the operating temperatures up to 220C, DELO® MONOPOX NT2999 or VK-26M adhesive manufactured by FSUE VIAM was used;

- the multilayer package is cut perpendicular to the size of 86.5 mm to achieve the final geometric size of 2.60x21.33x30.00 mm, which eliminates chipping of the edges of the magnets in the places of gluing. Next, grinding and polishing to size is carried out with a control measurement with a micrometer;

- the surface of the resulting magnet, having the final size, is subjected to phosphating, and then a protective coating of epoxy resin with a thickness of 10 to 35 microns is applied by spraying, providing a hardness of H6; the quality and thickness of the applied coating is controlled by X-ray fluorescence analyzer BA-100 (Bowman Analytics Inc);

- carry out the measurement of the main magnetic characteristics, including measurements of the magnetic dipole moment, as well as SST, PCT and HAST - tests. The magnetic dipole moment of finished SPMs is an integral characteristic of simultaneously volumetric (mass) and magnetic properties of SPMs. A comparative study of the dipole moments of 10 pieces of SPM using Magnet-Physik /Flux meter EP5/coil:M210 (Germany) at room temperature, as well as repeated measurements after heating for 1 hour on a metal sheet of 0.3 mm at a temperature of 200C, confirmed the deviation of the SPM dipole moments within 1.3%. The coating passed the SST test for 480 hours and the PCT test for 96 hours, the maximum working temperature of the coating was 200C; - carry out a comparison of the obtained magnetic characteristics with the parameters obtained as a result of numerical simulation and, with a positive result - the compliance of the SPM with the required characteristics - testing an experimental batch of permanent magnets as part of a mock-up and a prototype of the final product (electric generator) according to an agreed methodology and program.

Tests confirmed the possibility of processing raw materials of this subgroup A by the above "first" method. It was decided to process all raw materials of subgroup "A" according to this method. The rest of the raw materials (cuttings, chips, etc.) were sent for processing by hydrometallurgical methods.

Example 2. Raw materials of subgroup "B" were selected based on the results of the study of characteristics according to the methods described above.

As follows from Table 1, the raw materials of these manufacturers W-I-T had underestimated values of the coercive force Hcj (1-T) and overestimated values of aBr (W-I-T), which is apparently due to the fact that manufacturers did not add dysprosium in the manufacture of magnets in order to reduce the cost. This leads to a sharp increase in the temperature coefficient of induction, a sharp decrease in the value of W at operating temperatures and the emergence of defective batches of products.

The technology, in contrast to that described in example 1, is supplemented by the operation of local surface diffusion of dysprosium. After the operation of cutting plates to size on an EDM wire-cutting machine, followed by cleaning the surface of the resulting plates from EDM cutting products, the plates were subjected to local surface diffusion by applying a metal powder containing dysprosium, followed by heating. Such a procedure is described in the prior art (RU 2423204 C2). The exact weight percentage of dysprosium in diffusion was selected so that the temperature coefficient Br of each manufacturer's W, I, and T permanent magnets provided the required performance characteristics of the final product. As can be seen from Table 1, raw materials from these manufacturers had overestimated aBr values from -0.13 to -0.131%/°C. Dysprosium diffusion operation performed (1.5%) made it possible to increase the magnitude of the coercive force in terms of magnetization Hcj to 2800 kA/m at Т=20С and to reduce the value of аВг to -0.12%/°С, which corresponds to GOST R 52956-2008. After conducting diffusion and re-measurement of the magnetic properties, subsequent operations for the manufacture of SPM were carried out according to the method described in example 1.

Example 3 Raw material - about 1000 substandard NdFeB coated NdFeB magnets. REMs were rejected at the input control as having chips that exceeded the allowable dimensions formed during grinding. All REMs have magnetic properties close to the required ones, produced by one manufacturer from one raw material in one year with identical chemical composition.

An external examination established that the raw material was stored under normal conditions and there are no signs of corrosion. A random sample of rare-earth metals (approx. 0.1% of the batch) is made to clarify the magnetic characteristics, microstructure and other parameters. The obtained data for N42H REM at temperatures of 20 and 120C are presented in Table 3 (see the top two rows). It has been established that the raw material has magnetic characteristics at the lower boundary, while low values of Bg will not provide the required voltage at the output of the designed device.

Table 3

Characteristics for REM brand N42H at temperatures of 20 and 120C

Numerical modeling of a highly efficient electric generator with a power of 1 MW with an efficiency of 98.5%, in which REM will be operated at a temperature not exceeding 60C in peak mode, has been carried out. Such a low operating temperature of REM (low heat release due to Foucault currents) can be achieved by segmenting REM - gluing electrically insulated plates 0.5 mm thick into a multilayer package and placing the SPM along the axis of the electric generator. An additional numerical simulation is being carried out, in which 50% of the plates made of the N42) 1. material are replaced by plates made of the N50M material, which has a higher Br value. At this stage of numerical simulation, tabular parameters of the N50M brand are used according to the website (htp://www.amtc.ru/upload/ndfeb.pdf). It is shown that in the case under consideration the required performance characteristics of the electric generator can be achieved.

For the purpose of creating SPMs, N50M vacuum-packed REMs with a protective phosphate coating 1 µm thick were selected. They have a working temperature up to 100C, i.e. have a sufficient margin for operating temperature, ensuring the operability of the rotor of the electric generator in case of overheating.

Upon receipt of a batch of REM grade N50M from the manufacturer, a random sample is made to study the magnetic parameters, microstructure and other parameters. A random sample is made in the amount of not more than 0.1% of the batch. The results of measurements of magnetic characteristics at temperatures of 20 and 120C are presented in Table 3.

The electric generator is re-simulated with the parameters obtained as a result of measurements. It is shown that the gluing of alternating plates 0.5 mm thick from grades N42H and N50M in the SPM of final dimensions provides the required electrical parameters of the generator, including less than 1% voltage ripple. An electric machine is simulated with the following parameters: Вг=1.3408Тl, Hcj=1260kA/M at 20С, as well as with the obtained values of temperature coefficients.

To check the calculations, cylinders are cut from the multilayer package that forms the PSD for the study of magnetic characteristics and the HAST test. The measurement results are presented in Table 3.

The implemented technology, in contrast to example 1, is supplemented with operations for selecting the layout and orientation of single parts with different magnetic characteristics from two materials N42H and N50M in a formed multilayer package from the condition of obtaining the required integral magnetic characteristics. The sequence of technological operations of cutting and subsequent preparation of the surface of the obtained parts before gluing into the final SIM is preserved.

A prototype SPM was made, confirming the simulation data. The SPM consists of plates 0.5 mm thick and has geometric dimensions of 25.0 x 15.0 x 50.0 mm. The topology of the multilayer package is shown in Fig. 4, and the results of the magnetic studies are shown in FIG. 6,7,8. Fig. 6 shows the characteristics of NdFeB magnets N42H, Fig. 7 NdFeB magnets N50M, Figs. 7 - the characteristic of the SPM, formed by the characteristics of the SPM from alternating parts 20,21 of grades N42H and N50M, obtained at a temperature of 20C. To check the calculations, cylinders are cut from the multilayer package that forms the PSD for the study of magnetic characteristics and the HAST test. The measurement results are presented in Table 3.

EXAMPLE 4 Raw material - about 1000 substandard coated NdFeB grade N42H magnets. REMs were rejected at the input control as having chips that exceeded the allowable dimensions formed during grinding. All REMs have magnetic properties close to rowed ones and are produced by the same manufacturer.

The task is to manufacture an SPM that allows multiple overheating up to temperatures of about 300C. Simulation of an electric generator with the parameters obtained as a result of measurements is carried out. It is shown that gluing alternating plates 0.5 mm thick from NdFeB and SmCo grades into a SPM of final dimensions provides the required electrical parameters of the generator. To check the calculations, cylinders are cut from the multilayer package that forms the PSD for the study of magnetic characteristics and the HAST test. The implemented technology, in contrast to example 3, is supplemented with operations for selecting the layout and orientation of single parts with different magnetic characteristics from two materials of the NdFeB and SmCo brands in a formed multilayer package from the condition of obtaining the required integral magnetic characteristics. The sequence of technological operations of cutting and subsequent surface preparation of the obtained parts before gluing into the final SPM is preserved.

A prototype SPM was made, confirming the simulation data. SPM consists of plates with a thickness of 0.5 mm and has geometric dimensions 25.0 x 15.0 x 50.0 mm. Technological operations for the preparation of raw materials are similar to the operations of example 3. The simulation of the electric generator is carried out with the parameters obtained as a result of measurements. It is shown that gluing alternating plates 0.5 mm thick from NdFeB and SmCo grades in final sizes of SPM provides the required electrical parameters of the generator, including multiple overheating up to temperatures of about 300C. To check the calculations, cylinders are cut from the multilayer package that forms the PSD for the study of magnetic characteristics and the HAST test. The implemented technology, in contrast to example 3, is supplemented with operations for selecting the layout and orientation of single parts with different magnetic characteristics from two materials NdFeB and SmCo in a formed multilayer package from the condition of obtaining the required integral magnetic characteristics. The sequence of technological operations of cutting and subsequent surface preparation of the obtained parts before gluing into the final SPM is preserved. Measurements have shown that the SPM of the final dimensions, which is a package of glued alternating plates 0.5 mm thick from NdFeB and SmCo grades, can be operated up to temperatures of about 300C.

Industrial Applicability

The presented data confirm the achievement of the technical result - increasing the efficiency, completeness of processing of secondary magnetically hard sintered raw materials and expanding the range of semi-finished products, including: defective magnetic parameters, dimensions, chips, appearance and protective coating of elements of various magnetic systems, as well as previously used magnets, for the manufacture of effective segmented permanent magnets and magnetic systems with specified magnetic and geometric characteristics.

Claims

26 CLAIMS

1. The method of processing substandard magnetically hard sintered raw materials for the manufacture of segmented permanent magnets (SIM) includes the following operations: primary sorting of samples of raw materials into groups according to geometric ' - g characteristics and appearance, taking into account the required parameters of SIM and the allocation of raw materials for subsequent metallurgical processing; selective measurement of magnetic, chemical and microstructural parameters and carrying out a HAST test for each of the sorted groups of raw materials with the determination of the fundamental possibility of manufacturing SPM; selection of the first technology and production of an experimental batch of SPM from each sorted group, including cleaning the surface of samples, cutting samples into parts of a given configuration, gluing parts into a multilayer package, cutting and grinding the surface of the resulting multilayer package to a given size and applying a protective coating to the surface of the obtained SPM; measurement of the integral magnetic characteristics of the obtained SPM and comparison with the required characteristics; if the SPM does not meet the required characteristics, the secondary sorting of raw material samples is carried out into subgroups according to magnetic, chemical, microstructural parameters and the results of the HAST test for the selected second technology, including cleaning the surface of the samples, cutting the samples into parts of a given configuration, selecting the layout and orientation of single parts with different magnetic characteristics in the formed multilayer package from the condition of obtaining the required integral magnetic characteristics, gluing the parts into a multilayer package, cutting and polishing the surface of the resulting lot layer package to fit the size and applying a protective coating on the surface of the obtained SPM.

SUBSTITUTE SHEET (RULE 26)

2. The method according to claim 1, in which the substandard magnetically hard sintered neodymium-iron-boron and samarium-cobalt raw materials include magnetic elements that are rejected in terms of magnetic parameters, dimensions, chips, appearance and protective coating, as well as magnetic elements suitable for magnetic characteristics, but having a different size and previously used permanent magnets.

3. The method according to claim 1, in which the raw material for subsequent metallurgical processing, including hydrometallurgy methods, includes waste from cutting and grinding parts and a glued multilayer package, as well as permanent magnet breakage.

4. The method according to claim 1, in which the magnetic parameters include, but are not limited to, the parameters of the demagnetization curve at temperatures T = +20 and +150C: coercive force in induction Hcb and magnetization Hcj, residual induction Br, maximum energy product (BH)max, operating temperature range, Curie temperature Tc, as well as temperature coefficients a of reversible changes in the values aBr, aHcb, aHcj, as well as the direction of the axis of the magnetic texture, flux linkage, magnetic dipole moment, the nature of the distribution of the magnetic field induction created in the magnetic system of the final product.

5. The method according to claim 1, in which single parts with different characteristics of residual induction are made from raw materials of different grades, while the number of grades of raw materials included in one SPM is from 1 to 5.

6. The method of claim 1, wherein after cutting the specimens, the parts are subjected to a local surface diffusion procedure of dysprosium or terbium.

7. The method according to claim 1, in which the cutting of samples into parts of a given configuration, as well as a multilayer SPM package to size, is carried out by an electric spark or hydroabrasive method and / or a diamond disk, followed by mechanical and / or chemical elimination of the products of the specified machining.

SUBSTITUTE SHEET (RULE 26)

8. The method according to claim 1, in which cutting is carried out sequentially from the condition of minimizing chips in the sawing zone and creating the texture required for magnetization in the SPM, while first cutting the samples connected to each other, then cutting the obtained parts of a given size for gluing them into a multilayer package, and at the final stage - cutting the SPM with the required direction of the magnetic texture axis.

9. The method according to claim 1, in which the specified configuration of parts is selected from a list including, but not exhaustive: plate, parallelepiped, trapezoidal plate, cylinder, part of a hollow cylinder, including for Halbach cylinders.

10. The method according to claim 1, in which the gluing of the parts into a multilayer package is carried out with heat-resistant glue, while the glue is applied to both surfaces to be glued twice with intermediate drying, followed by joining the parts into a package and keeping the package under load until the adhesive cures.

11. The method according to claim 1, in which a protective coating with a hardness of H6 on the surface of a multilayer package of SPM is made of an epoxy compound with a thickness of 10 to 35 microns.

SUBSTITUTE SHEET (RULE 26)