JP4631298B2 - Rotating machine design method and rotating machine manufacturing method - Google Patents

Description

  The present invention relates to the design of a rotating machine that uses an electromagnetic action such as an electric motor or a generator directed at high efficiency and high accuracy.

  Rotating machines that use electromagnetic action, such as electric motors and generators, can be said to be indispensable for constituting today's automated society. For this reason, a more efficient rotating machine is desired because of the recent need to reduce energy loss. Moreover, in order to perform highly accurate control, it is necessary to implement | achieve a torque with high precision. For this reason, in the optimization design of rotating machines, minimizing loss and highly accurate prediction of torque are becoming increasingly important.

Conventional design methods for rotating machines are generally based on rough theory and experience. However, with such a design method, it is impossible to predict a specific numerical value with high accuracy even though a rough tendency of the device can be found from an analytical formula based on electromagnetism. Final judgment required trial and error by experience and experiment.
On the other hand, in recent years, with the improvement of calculation capability, simulation calculation (computer simulation) of a rotating machine using electromagnetic field analysis by a finite element method or an integral element method has become possible and is being applied to the design of equipment (non- (See Patent Documents 1 and 2). When optimization design is performed using these methods, a rough specification is determined from an analytical expression related to a rotating machine that has been used in the past, and the above specifications are performed while performing final rigorous calculation by electromagnetic field analysis. A general method is to determine the final specification by changing the contents.

Here, in the calculation by electromagnetic field analysis, the method of obtaining the optimum specification by comprehensively changing the design variable has a large calculation load, so that the design variable is asymptotic to the optimum solution as in Non-Patent Document 1. And a method for predicting an optimal solution from a limited calculation result based on an experimental design method, as in Non-Patent Document 2, has been proposed.
In any conventional design method, the material characteristics of the iron core used in the optimized design are the values measured directly from the iron core material itself, or the iron core material itself obtained from theory and experience. Iron core material properties are used.
Hino et al., “Design study of high-efficiency motor by iron loss analysis and shape optimization”, IEEJ rotating machine research data RM02-7, 2002 Fujishima et al., “Fundamental study on application of response surface approximation method to optimization calculation”, IEEJ rotating machine research data RM02-6, 2002

As described above, in the past, the optimum design specifications were obtained only by calculation on the computer, the prototype was actually manufactured using the obtained design, and the product characteristics were determined after determining the actual device characteristics. Transition to the production process.
In the design method using magnetic field analysis, device prediction can be performed with higher accuracy than the design method based on experience, but there may still be a discrepancy between the predicted value and the actual value. For this reason, there is a possibility that problems such as re-design change and re-design due to failure to obtain desired device characteristics may occur.

The reason for this divergence is that it is difficult for current electromagnetic field analysis to fully reflect the actual magnetic behavior such as anisotropy and hysteresis characteristics of iron core materials, and all the material characteristics of iron cores. There are various causes such as difficulty in quantifying strictly and the influence of strain introduced into the iron core material during motor manufacture.
In this way, the current simulation calculation technology does not take into account all of the various realistic factors, and must be performed under ideal and simplified conditions. There is a problem that the optimum solution is not always derived when applied to the optimization design of a rotating machine.
The present invention pays attention to such a point, and has an object to enable optimization design of a rotating machine with higher accuracy.

The inventors have found a phenomenon that cannot be reproduced by simulation (computer simulation) by measuring the local magnetic characteristics inside the stator of the motor, particularly by directly using the probe method. FIG. 4 to FIG. 6 show the comparison between the measured value by the probe method and the calculated value by the FEM (finite element method) for the locus of the local magnetic flux density vector, the local iron loss distribution, and the local magnetic flux density distribution inside the motor stator. Indicates. As is clear from these figures,
(1) Rotating magnetic flux behavior in the iron core that cannot be predicted by calculation (2) Deviation from the calculation result of local iron loss distribution (3) At least the phenomenon of deviation between the magnetic flux density inside the stator and the calculation of the distribution occurs I found out.
These are factors that cause an error in the result of the simulation calculation by the computer in designing the motor. The inventors considered that it is possible to perform highly accurate design calculation by performing complementation corresponding to each factor.

  Here, the rotating magnetic flux of (l) corresponds to the case where the magnetic flux density in the plane of the steel sheet constituting the iron core changes while having a component other than one direction, and rotation that causes periodic magnetization. In this machine, the origin of the magnetic flux density vector is the origin, and when the locus of the tip of the vector is drawn over one excitation cycle, the shape is drawn by rotating it into a circle or ellipse, or the shape in which high-frequency components are superimposed. Therefore, it is named in this way. It has been conventionally known that rotating magnetic flux is generated at the part where the direction of magnetic flux changes, such as the joint between the teeth and the yoke of the rotating machine and the tip of the tooth. However, in the past, no change in the direction of the magnetic flux was expected. The inventors have confirmed by local magnetic measurement that a rotating magnetic flux is generated for some reason even in a portion such as the central portion of the teeth. It is known that the iron loss is larger under the rotating magnetic flux than the alternating magnetic flux change (the iron loss under the rotating magnetic flux is called the rotating iron loss). Therefore, if the rotating magnetic flux behavior cannot be accurately grasped, the iron loss of the rotating machine cannot be accurately predicted, and an error occurs in the optimization that maximizes the efficiency. In addition, the rotating magnetic flux behavior, which is strongly related to the anisotropy of the iron core material, is thought to affect the torque pulsation of the motor. Therefore, if the rotating magnetic flux behavior is not accurately estimated, the torque pulsation However, it cannot be predicted accurately.

In particular, when a material with large magnetic anisotropy, such as a unidirectional electrical steel sheet or a bi-directional electrical steel sheet, is used as the iron core of a rotating machine, the magnetic anisotropy becomes stronger. Accurate reflection in electromagnetic field analysis is important for obtaining highly accurate results.
Conventionally, various methods for measuring the rotating magnetic flux and the rotating iron loss of the iron core material itself have been studied, and it has been studied to reflect this result in the electromagnetic field analysis of the rotating machine. For example, it is difficult to sequentially consider the influence of, for example, machining strain, which occurs in each part of the iron core in a damaged state. In addition, since it is difficult to create a strict database of rotating magnetic flux characteristics of iron core materials with consideration to hysteresis characteristics, for example, if the hysteresis characteristics affect the rotating magnetic flux behavior via the shape of the iron core, etc., the calculation accuracy will be reduced. It becomes a factor to make.

In this way, measuring the rotating magnetic flux behavior of the electrical steel sheet itself at the material level and reflecting this in the electromagnetic field analysis of the rotating machine is an optimal design in principle, but this is strictly At present, no method has been established for high-speed implementation.
In contrast, the inventors of the present application prototyped a primary model of the target rotating machine, locally measured the magnetic flux density inside the iron core of this rotating machine, in particular the behavior of the rotating magnetic flux, and calculated the electromagnetic field analysis. The present invention has been reached based on the idea that the characteristics of the actual rotating machine can be accurately calculated by changing / correcting the core material characteristic data so that the actual measured value approaches the actual measurement.

In other words, the present invention corrects the contribution of material properties such as anisotropy and hysteresis, which has been difficult with conventional general computer simulation (simulation calculation), by using experimental values, so that a highly accurate rotating machine design can be achieved. It provides iron core material characteristics that can realize the above.
Further, in today's design, a considerable part of the loss of the rotating machine is occupied by the loss (iron loss) generated in the iron core, and thus the above correction is made by paying attention to the iron core portion.

That is, the invention described in claim 1 of the present invention is a design method for a rotating machine that performs an optimized design of the rotating machine based on the core material characteristics and design specifications,
First, the magnetic properties of the core part of the rotating machine are calculated based on the predetermined iron core material characteristics and design specifications, and the test rotating machine manufactured with the core material and design specifications described above is driven . Measure the local magnetic properties of the iron core, correct the variables in the iron core material properties based on the amount of deviation between the calculated values and the measured values, and optimize the iron core material properties after the correction Determined as the core material characteristics used in the design ,
Next, in the optimization design, the specified core material characteristics are used, and the calculated characteristic value of the rotating machine while changing the design specifications other than the core material characteristics is designed to satisfy the target characteristics. It is what.

Next, in the invention described in claim 2, when the core material characteristics used in the optimization design are determined with respect to the configuration described in claim 1, the iron core material is used until the deviation amount becomes a predetermined value or less. The correction of the characteristics and the calculation of the magnetic characteristics of the iron core are repeated.
Next, the invention described in claim 3 uses an iron core material having the same material characteristics, and a plurality of design specifications for each design specification by the method described in claim 1 or claim 2, respectively. By specifying the core material characteristics of the above, a set of each of the above design specifications and the corrected core material characteristics is obtained in advance,
Based on the set of the iron core material characteristics and the corrected each design specification above previously obtained, defining a core material characteristics after the correction in the same or close to the design specifications and the rotating machine design specifications to produce,
Next, in the optimization design , the specified core material characteristics are used, and the calculated characteristic value of the rotating machine while changing the design specifications other than the core material characteristics is designed to satisfy the target characteristics. It is what.

Next, the invention described in claim 4 is characterized in that, for the configuration described in any one of claims 1 to 3, the design specifications are at least an iron core specification and a drive condition specification. To do.
Next, in the invention described in claim 5, with respect to the configuration described in any one of claims 1 to 4, the variable to be corrected among the core material characteristics relates to the magnetization curve of the core material. It is characterized in that it is at least one of a variable and a variable related to iron loss characteristics of the iron core material.
Next, in the invention described in claim 6, in contrast to the configuration described in any one of claims 1 to 5, the measurement of the magnetic characteristics of the iron core portion is performed by measuring the local magnetic flux density by the probe method. The measurement is performed by measurement.

In the following, the invention described in claim 7, intended to provide a method of manufacturing a rotating machine, characterized in that to produce using the design method according to any one of the above claims 1 to 6 is there.

  According to the present invention, among the specifications of the motor, the local data measured in a state close to the actual machine state, measured with a test machine, about the core material properties that are difficult to handle accurately by simulation calculation. By correcting based on the above, it is possible to accurately simulate the local magnetic behavior inside the actual iron core in a state incorporated in the rotating machine. By using the corrected iron core material characteristics, it becomes possible to perform the optimization design of the actual rotating machine with higher accuracy without making the calculation amount of the simulation calculation in the optimization design enormous.

  In addition, as described above, there are actually variables that affect the actual specifications in addition to the variables to be considered in the simulation calculation. In the present invention, the iron core material after the correction is also used for the variable amount. Is absorbed in the characteristics. In other words, the corrected iron core material characteristic is a value obtained by taking in other variable quantities that are not considered in the calculation with respect to the actual iron core material characteristic. Therefore, the iron core material characteristic after the correction is particularly reliable in a design specification that is the same as or close to the design specification used when obtaining the corrected iron core material characteristic.

Further, by measuring the magnetic flux density by the probe method, the magnetic flux density can be easily measured without adding processing to the testing machine.
Here, the important point in the present invention is that the corrected material magnetism suitable for calculating the local magnetic property inside the iron core in the state of being incorporated in the rotating machine from the measurement data and the calculation result of the magnetic property of the local region. It is to get characteristics. That is, such optimization can be performed based on a guideline that minimizes the average value of the deviation between the calculation result and the actual measurement result at each position of the iron core, for example. Alternatively, the inside of the iron core can be divided into several parts to minimize the difference between calculation and measurement. The method of correcting the calculation result by dividing the inside of the iron core is effective when, for example, the processing strain differs in each part of the iron core. Similarly, regarding the iron loss in the local region, the relationship between the magnetic flux density and the iron loss of the material can be changed for the entire iron core or for each part inside the iron core so as to minimize the deviation from the calculation result.

  According to the present invention, by correcting the iron core material characteristics based on the amount of deviation between the measured value and the calculated value in the magnetic characteristics, a material suitable for accurately calculating the local magnetic characteristics inside the iron core of an actual rotating machine with high accuracy. Magnetic characteristics can be obtained. Then, by performing the optimized design of the rotating machine using the corrected iron core material characteristics obtained, it becomes possible to obtain the optimized design specification of the rotating machine with a shorter time and with higher accuracy. .

Embodiments of the present invention will be described below with reference to the drawings.
FIG. 1 is a diagram showing an outline of the design method of the present embodiment, and is divided into two stages: a core material characteristic characteristic stage, which is the first characteristic part of the present application, and an actual rotary machine optimization design stage. The design stage is largely divided.
First, the processing procedure at the specific stage of the core material characteristics, which is the main feature of the present invention, will be described with reference to FIG.

In step S1, the core material characteristics and design specifications of the prototype are determined.
The prototype here is an actual prototype rotary machine for measuring local magnetic properties in the iron core. Details of the design specifications include, for example, drive conditions, design specifications of the iron core (size, ratio, etc. of each part). As a method for determining the specifications of the prototype, various methods such as a method using the existing specifications as they are, a method using a conventional analytical expression, and a method using electromagnetic field analysis can be employed. Here, the specification of the prototype is ideal as it is closer to the target optimum solution (optimized design). However, since the present invention performs the calculation for the optimization design using the more accurate core material characteristics incorporating the correction based on the actual measurement result, the calculation is not necessarily performed strictly at the prototype design stage. There is no need.

In step S2, the local magnetic characteristics inside the iron core when the prototype with the determined specifications is driven are calculated by simulation calculation using electromagnetic field analysis such as FEM.
On the other hand, in step S3, an actual rotating machine is manufactured based on the specifications determined in step S12. Subsequently, in step S4, the interior of the iron core when driven under actual driving conditions or conditions close thereto is used. Measure the local magnetic property distribution. The measurement here is preferably performed over the entire area of the iron core as much as possible. However, if the measurement region is limited, only a representative local portion of the iron core may be used. Further, in the present invention in which components that cannot be accurately reproduced by calculation are corrected by actual measurement, the local magnetic measurement here is performed in the r direction (radial direction of the rotating machine) and the θ direction (circumferential direction of the rotating machine). The accuracy of correction can be increased by using the two-dimensional local magnetic measurement performed in two directions.

  Here, it is preferable to manufacture a rotating machine at each trial production from the viewpoint of improving the design accuracy, but the measured value of the local magnetic characteristics measured by making an actual rotating machine is measured. As a result, the design period can be greatly shortened and the accuracy can be improved more than before by using the one selected from the measured values of the multiple local magnetic characteristics measured for rotating machines manufactured with multiple design specifications. Design becomes possible.

Subsequently, in step S5, the calculation result of the previous electromagnetic field analysis and the result of local magnetic measurement are compared to determine the amount of divergence between them. In step S6, the amount of divergence is a predetermined target set in advance. It is evaluated whether or not the value is equal to or less than the value. If the value is equal to or less than the predetermined target value, the process proceeds to the optimization design stage. If the deviation is larger than the predetermined target value, the process proceeds to step S7.
Here, as a target of the deviation amount to be evaluated, there are local magnetic flux density (magnetic flux density in a local portion inside the stator), local iron loss characteristics (iron loss in a local portion inside the stator), and the like. Conceivable.
The divergence amount target amount is set to, for example, several percent of the initial value of the magnetic characteristic value that is the divergence amount target amount, according to the required design accuracy.

In step S7, based on the deviation amount obtained in step S5, the values of variables related to the magnetization curve, iron loss, and the like of the iron core material are changed in a direction in which the deviation amount is considered to be small, so that the iron core material is changed. Correct the characteristics. Then, it transfers to said step S2, performs electromagnetic field analysis (calculation by a computer) again using the changed iron core material characteristic, and compares with an actual measurement result (S5, S6).
Here, as the variable to be changed, for example, with respect to the magnetic permeability, for example, each component of the magnetic permeability matrix of the following equation (1) and the angle θ between the H vector and the B vector can be cited. And the coefficient and function form which describe these are changed. It is possible to change the coefficients (c ₁ , c ₂ ) of the formula (2) relating to the iron loss. In equations (1) and (2), Br and Hr are the magnetic flux density and the magnetic field diameter component, respectively, and Bθ and Hθ are the magnetic flux density and the magnetic field circumferential component, respectively.

Here, the matrix component μ is a function of the magnetic field H, θ is the angle between the H vector and the B vector, and the above equation is a function of the magnetic field H and the frequency. The μ and θ are variables related to the magnetization curve and iron loss, and are changed according to the amount of deviation.
W = c ₁ · f · Bm ^1.6 + c ₂ · f ² · Bm ² (2)
Here, c ₁ and c ₂ are coefficients Bm are the maximum values of the magnetic flux density vector B, and f is the frequency.

According to the above-described processing at the specific stage of the core material characteristics, in the normal electromagnetic field analysis and the optimization calculation using the electromagnetic field analysis, the permeability μ, the coefficient c ₁ , While c ₂ is used as it is, the value measured with the iron core material, it is corrected to match with the actual measurement result in the local region of the iron core, so that the actual value in the state incorporated in the rotating machine. As a result, it is possible to substantially improve the calculation accuracy in the next optimization design stage.
The change of the material variable for bringing the calculation result closer to the actual measurement result may be performed based on some standard, or the optimum solution may be searched for while randomly changing the numerical value. Alternatively, in order to obtain an optimal solution, the number of repeated calculations and the difference between the target calculation and the measured value can be arbitrarily selected as necessary.

As a material variable optimization method, for example, the following procedure can be considered.
First, the magnetization curve of the material _{B = a 1 / [a 2} + (a 3 H + a 4) a5 + a 6 H + a 7
: When a ₁ ~a ₇ represents a function such as coefficient determined by the material, the magnetic permeability μ μ = μ (H) = (a 1 / [a 2 + (a 3 H + a 4) a5 + a 6 H + a 7) / H It becomes.

Next, for the formula (1), coefficients c ₁₁ , c ₁₂ , c ₂₁ , c ₂₂ for fitting are introduced,
μ ₁₁ = c ₁₁ μ (H)
μ ₁₂ = c ₁₂
μ ₂₁ = c ₂₁
Let μ ₂₂ = c ₂₂ μ (H).
Let B ^calc = (Br ^calc , Bθ ^calc ) be the local magnetic flux density vector locus calculated by the magnetic field analysis of the motor of the initial model using the coefficient defined above and θ of equation (1). When the measured local magnetic flux density vector locus is B ^meas = (Br ^meas , Bθ ^meas ), the average value of the deviation between the measured value and the calculated value over one cycle is

And
Here, ω is the rotation angle of the local magnetic flux density vector, and the integration according to the above equation is performed in the range of 0 to 2π.
Here, the average value of ΔB at the measurement position numbers 1 to N is <ΔB>,
Further, the allowable value of <ΔB> is set to 3% of the average value <Bm> of the maximum values of the magnetic flux density at the measurement point numbers 1 to N.
That is, <ΔB> / <Bm><0.03 is set as the target condition.

Next, the coefficients c ₁₁ , c ₁₂ , c ₂₁ , c ₂₂ and θ are slightly increased or decreased manually or automatically from the initial values of c ₁₁ = c ₁₂ = 1, c ₁₂ = c ₂₁ = 0, respectively. <ΔB> / <Bm> is sequentially calculated to find a combination (c ₁₁ , c ₁₂ , c ₂₁ , c ₂₂ , θ) that satisfies the above target condition. As the initial value of θ, it is preferable to use θ obtained as a result of two-dimensional magnetic measurement of the material, and is usually about 25%.
By applying the permeability matrix obtained as described above to the electromagnetic analysis of the motor, and by optimizing the design of the rotating machine, it is possible to improve the motor characteristics by simulation calculation with higher accuracy than before. It becomes.

Next, the optimization stage process (optimization design) will be described.
The design method is the same as the conventional design method for optimization except that the optimized core material characteristic obtained by the above-described process, that is, the material variable is changed, is used. That is, using the corrected iron core material characteristics obtained above, optimization calculations (electromagnetic field analysis) are performed while changing variables such as the size and ratio of each part of the rotating machine other than the iron core material characteristics. A rotating machine design specification that achieves the predetermined target characteristic is obtained (S8 to S15).
Then, an actual rotating machine is manufactured based on the obtained design specifications and the evaluated core material.

  When a rotating machine is designed by the above processing, accurate results can be obtained for convenience even when using a magnetic characteristic curve that simplifies the complex magnetic behavior of an actual magnetic material. The machine can be optimized. That is, since it is possible to carry out computer simulation and optimization design of a rotating machine with higher accuracy than before, it is possible to obtain high rotating machine characteristics with high accuracy.

The method of the present invention is related to the magnetic properties of an electrical steel sheet that is difficult to handle due to its complexity in practice. In order to improve the accuracy of the calculation results for the sake of convenience, a correction is made to the simplified characteristic curve. Although there are unknown variable quantities that are not taken into account in the calculation, the optimization design of the rotating machine can be performed with high accuracy by incorporating the variable quantities into the core material characteristics. If the processing method of the prototype is the same as the processing method of the actual rotating machine, an unknown variable amount at the time of manufacturing such as processing strain is also taken into the core material characteristics.
Here, it is preferable to select the prototype specifications at the specific stage of the core material characteristics and the specifications of the actual rotating machine at the optimization stage that are the same or close. As described above, unknown variable quantities (such as errors due to machining strain) are also included in the core material characteristics, so that the reliability of the core material characteristics after correction improves as the specifications of both are closer.

  In the above description, every time optimization design is performed, a prototype for correcting the core material characteristics is manufactured. However, the present invention is not limited to this. For iron core materials made of the same iron core material, after performing a specific stage processing of the above iron core material characteristics in advance for each of a plurality of different design specifications (a plurality of design specifications with different core specifications or driving condition specifications) From the multiple design specifications, the design specifications close to the actual design specifications (that is, the difference from the target characteristic value of the desired rotating machine is less than the predetermined value) The optimized design specifications of the rotating machine may be obtained by appropriately selecting and using the corrected iron core material characteristics corresponding to the design specifications obtained by the rotating machine. In this case, the design period is greatly shortened.

Here, it is assumed that a plurality of core material characteristics determined in advance are selected, but when there is no design specification that is the same as the design specification of the actual motor, a plurality of corrections corresponding to two or more design specifications close to the actual design specification The core material characteristics of the corresponding design specification may be determined using the average value of the core material characteristics.
Or, based on a set of core material characteristics and design specifications obtained in advance, put the core material characteristics as a function expression or graph of the design specifications, and calculate the core material characteristics to be used using the function expression etc. It may be specified.

  The applicable range of the present invention is applicable to any rotating machine (electric motor, generator) based on electromagnetic action. Examples of the electric motor include a brushed DC motor, a brushless DC motor, an induction motor, a reluctance motor, and the like, and any of these can be applied. Of these, brushless DC motors have been gaining importance in recent years because of their high efficiency, and application of the present invention that contributes to improving motor efficiency is extremely useful. There are various types of generators, such as a type that uses an electromagnet for the field, a type that uses a permanent magnet, and a type that provides a switched reluctance motor with a power generation function, and any of them can be applied.

  Any conventional method such as an analytical method, finite element method, or integral element method can be applied to the calculation of rotating machine characteristics by a computer. In consideration of correction, the calculation method of the rotating machine characteristics by the computer needs to be able to calculate the magnetic characteristics of the local region. In addition to the non-linearity of the magnetic characteristics of the material, it is necessary that calculations such as two-dimensional characteristics and iron loss characteristics can be performed with a certain degree of accuracy.

  As described above, the prototype rotating machine used for the actual measurement of the local magnetic characteristics is a method using the existing specifications as it is, a method using a conventional analytical method, an electromagnetic field analysis, etc. Various methods can be employed, for example, in the case of using a conversion method. In addition, actual measurement results with several typical rotating machines are obtained before a specific design problem, and a new optimization design is performed using material properties corrected and changed based on this result. Also good. In this case, whether the data for a typical rotating machine obtained in advance can be applied to a new design depends on how close the specifications and required characteristics of the typical rotating machine and the new rotating machine are. However, compared with the conventional method, simulation calculation can be performed with higher accuracy.

The amount of comparison between the calculation result and the experimental result is preferably the behavior of the magnetic flux density in the local region or the iron loss. This is because these local quantities together form the torque and efficiency of the rotating machine. Moreover, the correction | amendment performed in order to make the deviation | shift amount regarding the said quantity below a predetermined value is good to change variables, such as a permeability matrix and an iron loss curve.
Also, when reducing the amount of deviation between the calculated value and the measured value, it is better to use the same conditions as the rotating machine for which the operating conditions such as the rotational speed and torque are designed, but the rotational speed and torque, etc. Is used in a wide range, it is preferable to evaluate the deviation amount using an average value or a weighted average value under the use conditions.

The change of the variable related to the core material characteristics may be performed using an average value over the entire core, or the inside of the core may be divided into several regions, and different variables may be given to each part.
In the measurement of the local magnetic flux density in the iron core, a conventionally used probe coil method may be used, or a probe method, a Rogowski coil method, or the like can be applied. Among these methods, the probe method is the most excellent method in that non-destructive and quick multipoint measurement is possible. In addition, since the two-dimensional rotating magnetic flux behavior cannot be accurately predicted by the conventional electromagnetic field analysis, it is preferable to actually measure the local magnetic flux density in two dimensions. FIG. 2 shows a schematic diagram relating to the local magnetic measurement inside the rotating machine iron core by the probe method. Japanese Unexamined Patent Publication No. 2000-352579 has already disclosed a method for measuring local magnetism inside a rotating machine stator by a probe method. In particular, this probe method guarantees the measurement accuracy of the magnetic flux density in a sufficiently thin magnetic material, but the effect is exhibited if the iron core is a type in which plate-like magnetic bodies are laminated.

  When evaluating the local iron loss, the magnetic flux density in the local region is measured by the probe method, and the magnetic field on the steel sheet surface is measured by placing a magnetic field detection sensor such as a Hall element at the center of the two probes. The method for obtaining the local iron loss from the magnetic flux density and the magnetic field in the local region is excellent in accuracy and speed. In the local iron loss measurement by this method, the iron loss components in two directions orthogonal to each other are measured in the same manner as the magnetic flux density measurement, and the local iron loss can be obtained by taking these sums. As described above, according to the measurement probe including the probe and the magnetic field detection element, it is possible to quickly measure the magnetic flux density, the magnetic field, and the iron loss in the local region, and the usefulness of the present invention can be enhanced. is there. As a method for measuring local iron loss, there is also a method for measuring a local temperature rise, which can be applied to the present invention.

The target value at the optimization stage of the rotating machine is important for the rotating machine such as motor torque, torque pulsation, and efficiency, and can be any numerical value that can be calculated by electromagnetic field analysis. In general, an optimum design can be obtained in a direction in which efficiency and torque are maximized and torque pulsation is minimized. The present invention is also useful for optimization that attempts to minimize the size of the rotating machine under conditions where torque, torque pulsation, efficiency, etc. are within a certain range.
The specifications changed for the optimization of the rotating machine design include items related to the iron core itself such as the size, shape, number of windings and connection method of the iron core, as well as the driving / control method of the rotating machine.

"Example 1"
In the 8-pole 12-slot concentrated winding surface magnet type brushless DC motor (output: 300 W, diameter: 160 mm) shown in FIG.
1) Using a JIS grade 35A300 non-oriented electrical steel sheet as the iron core material, a prototype motor was designed based on the past specifications of the same type of motor (iron core material characteristics and design specifications).
2) A simulation of the finite element method (FEM) by two-dimensional static magnetic field analysis was performed on the computer with respect to this motor specification, and the behavior of the magnetic flux density in the local region was calculated.
3) A test motor for measuring the local magnetic characteristics was prototyped according to the specifications of 1) above. At this time, the windings of the teeth were raised so that the magnetic flux density inside the teeth could be measured by the probe method.
4) The local magnetic flux density of the test motor of 3) above was measured two-dimensionally by the probe method.
5) B ^cal = (Br ^cal , Bθ ^cal ) for the magnetic flux density in the local region at the measurement point P (r, θ) [where r and θ are radial and circumferential directions], and B ^meas = ( Br ^meas , Bθ ^meas ), and when the difference ΔB between the calculated value and the measured value is expressed by the following equation:
ΔB = | {(Br ^cal ) ² + (Bθ ^cal ) ² } ^1/2
− {(Br ^meas ) ² + (Bθ ^meas ) ² } ^1/2 |

Calculation was performed by appropriately changing the variables (θ, μ ₁₁ , μ ₁₂ , μ ₂₁ , μ ₂₂ ) so that the average value <ΔB> of the deviation amount ΔB over all the measurement points was equal to or less than the deviation amount target value. Here, as a calculation method of these variables, when the magnetic permeability of the material is μ, θ 11 is first calculated by setting μ ₁₁ = μ ₂₂ = μ and μ ₁₂ = μ ₂₁ = 0, and the θ that matches the actual measurement result is obtained. and then, fixing the _{θ μ 11 = μ 22 = μ} , μ 12 = μ 21 = 0 while shifting a value slightly from <△ B> value less divergence amount target value (function) (mu _11, μ _22, μ _12, was determined mu _21).

Subsequently, an electromagnetic field analysis was performed on the computer by driving voltage waveform input using these variables: θ, μ ₁₁ , μ ₂₂ , μ ₁₂ , and μ ₂₁ , and the iron loss was calculated based on the above formula (2) from the result. The motor design was optimized to maximize motor efficiency. In the optimization calculation, the stator outer diameter was constant, and the variables shown in FIG. 3 (yoke width a, tooth width b, tooth tip angle c, gap length d, distance e between adjacent tooth tips) were changed.
The motor was manufactured according to the optimization design obtained by the optimization of the motor design. Table 1 shows the maximum efficiency measured using the manufactured motor and the conventional method and the test motor before optimization. Here, the conventional method is the optimum calculation obtained from the electromagnetic field analysis without any change or correction of the magnetic material characteristics.

As shown in Table 1, it can be seen that when optimization design is performed using the iron core material after correction based on the present invention, high efficiency that cannot be obtained by the conventional method is obtained.
"Example 2"
In the 8-pole 12-slot concentrated winding surface magnet type brushless DC motor (output 300 W, diameter 160 mm) shown in FIG. 3, the optimization design of the motor was performed by strictly evaluating the rotating magnetic flux behavior by the following procedure.

1) Using a JIS grade 35A300 non-oriented electrical steel sheet as the iron core material, a prototype motor was designed based on the past specifications of the same type of motor (iron core material characteristics and design specifications).
2) A simulation of the finite element method (FEM) by two-dimensional static magnetic field analysis was performed on the computer with respect to this motor specification, and the behavior of the magnetic flux density in the local region was calculated.
3) A test motor for measuring local magnetic properties was made on the basis of the specifications of 1) above. At this time, the winding of the tooth part was raised so that the magnetic flux density inside the tooth could be measured.
4) The local magnetic flux density of the test motor of 3) above was measured two-dimensionally by the probe method.
5) B ^cal = (Br ^cal , Bθ ^cal ) for the magnetic flux density in the local region at the measurement point P (r, θ) [where r and θ are radial and circumferential directions], and B ^meas = ( Br ^meas , Bθ ^meas ), and when the difference ΔB between the calculated value and the measured value is expressed by the following equation:
ΔB = | {(Br ^cal ) ² + (Bθ ^cal ) ² } ^1/2
− {(Br ^meas ) ² + (Bθ ^meas ) ² } ^1/2 |

The variables (θ, μ ₁₁ , μ ₁₂ , μ ₂₁ , μ ₂₂ ) were calculated so that the average value <ΔB> of the deviation amount ΔB over all the measurement points was less than the deviation amount target value. Here, as a calculation method of these variables, when the magnetic permeability of the material is μ, θ 11 is first calculated by setting μ ₁₁ = μ ₂₂ = μ and μ ₁₂ = μ ₂₁ = 0, and the θ that matches the actual measurement result is obtained. and then, fixing the _{θ μ 11 = μ 22 = μ} , μ 12 = μ 21 = 0 while shifting a value slightly from <△ B> value less divergence amount target value (function) (mu _11, μ _22, μ _12, was determined mu _21).

Subsequently, the electromagnetic field analysis by the drive voltage waveform input using these variables: θ, μ ₁₁ , μ ₂₂ , μ ₁₂ , μ ₂₁ is performed on the computer, the motor efficiency and torque are calculated, and the motor efficiency is 92%. As described above, the motor design is optimized so as to minimize the torque pulsation under the condition of the torque of 2 Nm. Here, the torque pulsation amount was defined by the following equation (3).
Torque pulsation [%]
= {(Average value of absolute value of deviation from average torque) / average torque} × 100 (3)

In the optimization calculation, the stator outer diameter was constant, and the variables shown in FIG. 3 (yoke width a, tooth width b, tooth tip angle c, gap length d, distance e between adjacent tooth tips) were changed.
The motor was manufactured according to the optimization design obtained by the optimization of the motor design. Table 2 shows the torque pulsation amount measured using the manufactured motor and the torque pulsation amount measured by the conventional method and the test machine before the optimization calculation. Here, the conventional method is the optimum calculation obtained from the electromagnetic field analysis without any change or correction of the material magnetic properties.

As can be seen from Table 2, the accuracy of analysis is improved by applying the present invention, and thus a motor with less torque pulsation can be designed.
"Example 3"
In the 8-pole 48-slot distributed winding / embedded magnet type brushless DC motor (output 10 kW, diameter 200 mm), the optimization design of the motor was performed by strictly evaluating the rotating magnetic flux behavior according to the following procedure.

1) Using a non-oriented electrical steel sheet of JIS grade 35A250 as the iron core material, a prototype motor was designed based on the specifications of the same type of motor in the past.
2) A simulation calculation by a finite element method (FEM) was performed on the computer for this motor specification, and the behavior of the magnetic flux density in the local region was calculated.
3) A test motor for measuring the local magnetic characteristics was prototyped from the motor having the specifications of 1) above. At this time, the winding of the tooth part was raised so that the local iron loss inside the tooth could be measured.
4) The local iron loss of the test motor of 3) above was measured by a probe method. At this time, the rotational speed is changed in the range of 1000 to 5000 rpm in increments of 500 rpm, the torque is changed in the range of 20 to 50 Nm in increments of 10 Nm, and the motor efficiency is measured and probed at each rotational speed and torque condition. The local magnetism was measured with a probe consisting of a method and a Hall element.
5) The iron loss in the local region at the measurement point P (r, θ) [r and θ are radial and circumferential directions] was calculated by the following equation.
W = ∫ {Σ (c ₁ · f _i · B _i ^1.6 + c ₂ · f _i ² · B _i ² )} dv (4)

Here, f _i is the frequency of each harmonic component of the magnetic flux density waveform by FFT analysis, B _i is the peak value of each harmonic component of the magnetic flux density waveform, and Σ is the sum of harmonic components up to a sufficiently high order. , ∫ {} dv is the sum for the entire stator. c ₁ and c ₂ are coefficients depending on the material. In the case of an embedded magnet type motor, a loss is also generated inside the rotor as compared with a surface magnet type motor, but this is neglected here because it is smaller than the stator iron loss.

Subsequently, the difference between the calculated value of the local iron loss and the actually measured value is divided into (1) a tooth tip portion, (2) a teeth portion, (3) a joint portion between a tooth and a yoke, and (4) a yoke portion as follows ( 5) Obtained by the equation.
ΔW = | W ^cal −W ^meas | (5)
Where W ^cal : Iron loss by calculation based on equation (4)
W ^meas : Local iron loss measured using the local magnetic measurement probe of FIG.

Subsequently, an average value <ΔW> of ΔW is obtained for each of the above parts (l) to (4), and variables (c ₁ , c ₂ ) are set for each part so as to be equal to or less than the deviation amount target value. Asked. Here, as a calculation method of these variables, <ΔW> of each part of (1) to (4) is less than the deviation amount target value while gradually shifting the values of c ₁ and c _{2 in} the equation (4). Values (c ₁ , c ₂ ) were obtained.
Subsequently, the motor design was optimized by electromagnetic field analysis using c ₁ and c ₂ which are variables of the iron core material characteristics after correction. In the optimization calculation, the outer diameter of the stator was made constant, and the variables shown in FIG. 3 were changed (the motor specifications are different from those in FIG. 3 but the variable determination method is the same).
Table 3 shows the efficiency at the time of optimization design obtained by applying the present invention in this way and the error between the calculated value and the predicted value together with the result of the conventional method. Here, the conventional method is the optimum calculation obtained from the electromagnetic field analysis without any change or correction of the material magnetic properties.

As shown in Table 3, the motor designed by the method according to the present invention has a higher average efficiency than that obtained by the conventional optimization method.
Next, a second embodiment will be described with reference to the drawings.
In the present embodiment, an example of an iron core material characteristic specifying device 10 that calculates an optimized iron core material characteristic after correction will be described.
As shown in FIG. 7, the present apparatus receives as input data the predetermined core material characteristics and design specifications, and the measured magnetic characteristics of the core measured by the testing machine. Each input data is stored in the storage unit 10A. Note that the core material characteristics are set as formulas in the characteristic changing means 10D, which will be described later, and actually, variable values of formulas representing the core material characteristics are input to the apparatus. The same applies to the design specifications.

The iron core material property specifying device 10 includes a storage unit 10A, a magnetic property calculation unit 10B, a property change unit 10D, and a comparative evaluation unit 10C.
The magnetic characteristic calculation means 10B performs simulation calculation based on the well-known electromagnetic field analysis as described above using the core material characteristic and design specification data stored in the storage unit 10A. A calculated value for comparative evaluation is obtained and output to the comparative evaluation unit 10C.

  The evaluation comparison unit compares the calculated value input from the magnetic characteristic calculation means 10B with the measured magnetic characteristic, calculates the divergence amount between them, and determines whether the divergence amount is equal to or less than a predetermined divergence amount target value. If it is determined and the deviation amount is equal to or less than the target value, the iron core material characteristic stored in the storage unit 10A is output as an optimized iron core material characteristic after correction to a display unit or database (not shown). On the other hand, if the deviation amount is larger than a predetermined deviation amount target value, the deviation amount is output to the characteristic changing means 10D.

The characteristic changing means 10D changes a predetermined variable among the core material characteristics, stores it again in the storage unit 10A, and sends a recalculation command to the magnetic characteristic calculation means 10B. Note that the amount of change may be constant, or the amount of change may be determined according to the amount of deviation.
In the core material characteristic specifying device 10 having the above-described configuration, an optimized core material characteristic can be obtained only by inputting a predetermined core material characteristic and design specification and a measured magnetic characteristic of the core measured by a testing machine. It becomes possible.

Next, a third embodiment will be described with reference to the drawings.
In the present embodiment, an example of the core material property specifying device 12 that specifies the optimized core material property after correction will be described.
As shown in FIG. 8, this apparatus receives data and design specifications for specifying a predetermined core material, and outputs the optimized core material characteristics after correction with reference to the database 11.

Here, the database 11 is preliminarily stored with data including a table including at least each design specification data and the corrected core material characteristic data for each core material obtained by the apparatus or other methods as in the second embodiment. Stored.
And the said iron core material characteristic specific | specification apparatus 12 searches the table corresponding to the input iron core material from the database 11, and acquires the table corresponding to the data which is the same as or close | similar to the input design specification in the table.
Subsequently, if there is a table of design specification data that is the same as the input design specification, the corrected core material characteristics of the table are output.

On the other hand, if there is no table of design specification data that is the same as the input design specification, multiple tables of design specification data close to the input design specification are selected, and a predetermined calculation method (average) is selected from the core material characteristics in the selected table. The core material properties are calculated by complementing the values by taking values, and the calculated core material properties are output.
In the core material characteristic specifying device 12 having the above-described configuration, an optimized core material characteristic can be obtained only by inputting a predetermined core material characteristic and design specification, and a measured magnetic characteristic of the core measured by a testing machine. It becomes possible.
In addition, the characteristic period of the motor specification is shortened by the amount that the magnetic characteristic measured by making the test machine is not obtained for each optimum calculation.

  According to the present invention, it becomes possible to predict the characteristics of a rotating machine with higher accuracy than before, and by applying this to design optimization, it is possible to obtain a rotating machine that is superior in efficiency, torque characteristics, and the like. .

It is a figure explaining the design procedure which concerns on 1st Embodiment based on this invention. It is a figure which shows the state of the probe method which concerns on 1st Embodiment based on this invention. It is a figure which shows the iron core specification of the Example which concerns on 1st Embodiment based on this invention. It is a figure which shows the locus | trajectory of a local magnetic flux density vector. It is a figure which shows distribution of a local iron loss. It is a figure which shows distribution of local magnetic flux density. It is a block diagram which shows the example of the iron core material characteristic apparatus which concerns on 2nd Embodiment based on this invention. It is a block diagram which shows the example of the iron core material characteristic apparatus which concerns on 3rd Embodiment based on this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Iron core material characteristic identification apparatus 10A Memory | storage part 10B Magnetic characteristic calculation means 10C Comparison evaluation part 10D Characteristic change means 11 Database 12 Iron core material characteristic identification apparatus

Claims (7)

A design method for a rotating machine that performs an optimized design of the rotating machine based on iron core material characteristics and design specifications,
First, the magnetic properties of the core part of the rotating machine are calculated based on the predetermined iron core material characteristics and design specifications, and the test rotating machine manufactured with the core material and design specifications described above is driven . Measure the local magnetic properties of the iron core, correct the variables in the iron core material properties based on the amount of deviation between the calculated values and the measured values, and optimize the iron core material properties after the correction Determined as the core material characteristics used in the design ,
Next, in the optimization design, the specified core material characteristics are used, and the calculated characteristic value of the rotating machine while changing the design specifications other than the core material characteristics is designed to satisfy the target characteristics. Rotating machine design method . 2. When determining the core material characteristics to be used in the optimization design, the correction of the core material characteristics and the calculation of the magnetic characteristics of the core section are repeated until the deviation amount becomes a predetermined value or less. The design method of the rotating machine described in 1. By using the iron core material having the same material characteristics, for each of the plurality of design specifications, the core material characteristics for each design specification are specified by the method according to claim 1 or 2, respectively. Find a set of design specifications and corrected core material properties in advance,
Based on the set of the iron core material characteristics and the corrected each design specification above previously obtained, defining a core material characteristics after the correction in the same or close to the design specifications and the rotating machine design specifications to produce,
Next, in the optimization design , the specified core material characteristics are used, and the calculated characteristic value of the rotating machine while changing the design specifications other than the core material characteristics is designed to satisfy the target characteristics. Rotating machine design method . The said design specification is an iron core part specification and a drive condition specification at least , The design method of the rotating machine of any one of Claims 1-3 characterized by the above-mentioned. The variable to be corrected among the iron core material characteristics is at least one of a variable related to a magnetization curve of the iron core material and a variable related to iron loss characteristics of the iron core material. A method for designing a rotating machine as described in 1 above. The method for designing a rotating machine according to any one of claims 1 to 5, wherein the measurement of the local magnetic characteristics of the iron core is performed by local magnetic flux density measurement by a probe method. . It manufactures using the design method of any one of the said Claims 1-6, The manufacturing method of the rotary machine characterized by the above-mentioned.

Images