JP2016065773A - Rotary machine diagnostic device and rotary machine diagnostic method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve correctness of vibration diagnosis of a rotary machine.SOLUTION: A rotary machine diagnostic device of one embodiment comprises: a vibration application part for applying vibration force to a k-th part (k is an integer satisfying 1≤k≤n) out of first to n-th parts of a structure of the rotary machine (n is an integer of equal to or more than 2); a first vibration response measurement part for measuring a first vibration response which is a vibration response of the k-th part to the vibration force applied to the k-th part; and a second vibration response measurement part for measuring a second vibration response which is a vibration response of the first part to the vibration force applied to the k-th part. The device further comprises a transfer function calculation part for, based on the first vibration response, calculating a first transfer function which is a transfer function of the k-th part to the vibration force applied to the k-th part, and calculating a second transfer function which is a transfer function of the first part to the vibration force applied to the k-th part, and a diagnosis part for diagnosing whether or not, vibration intensity of the structure is allowable, based on the first and second transfer functions.SELECTED DRAWING: Figure 1

Description

  Embodiments described herein relate generally to a rotating machine diagnostic apparatus and a rotating machine diagnostic method.

  FIG. 8 is a schematic diagram showing a configuration of a general rotating machine diagnostic apparatus.

FIG. 8A and FIG. 8B show a stator 1 of a rotating machine, a coil end 2 of the stator 1, and a rotating machine diagnostic apparatus that diagnoses the rotating machine. An example of the rotating machine is a rotating electric machine such as a turbine generator. The coil end 2 includes first to nth coils C _{1 to} C _n (n is an integer of 2 or more). The rotating machine diagnostic apparatus includes a load cell 3, an impulse hammer 4, an acceleration pickup 5, a power unit 7, an FFT (Fast Fourier Transform) analyzer 8, and an arithmetic device 9.

  8A and 8B, the X direction and the Y direction indicate the horizontal direction, and the Z direction indicates the vertical direction. The X direction is parallel to the axial direction of the stator 1, and the Y direction is perpendicular to the axial direction of the stator 1. FIG. 8A shows a cross section perpendicular to the Y direction at the coil end 2. FIG. 8B shows a cross section perpendicular to the X direction at the coil end 2.

  The coil end 2 is a conical shell-like structure projecting outward in the axial direction of the stator 1 with respect to the stator (stator core) 1. The coil end 2 is formed by binding a plurality of coils. As shown in FIG. 8B, the coil end 2 has a periodically symmetric structure. An electromagnetic excitation force having a frequency twice as high as the power frequency of the rotating machine acts on the coil end 2.

  As shown in FIG. 8A, each coil of the coil end 2 has a cantilever structure and does not have high rigidity like the stator 1. Therefore, the coil end 2 is likely to vibrate due to the electromagnetic excitation force and is easily damaged due to the vibration. Therefore, the coil end 2 is inspected at the time of manufacture or inspection of the rotating machine, and the coil end 2 has sufficient vibration strength with respect to the electromagnetic excitation force, that is, the coil end 2 has the electromagnetic excitation force. It is desirable to confirm that it does not vibrate very much.

  The rotating machine diagnostic apparatus is used for diagnosing whether or not the vibration intensity of the coil end 2 is acceptable. Specifically, whether or not the coil end 2 has sufficient vibration strength with respect to a predetermined excitation force, that is, whether or not the vibration of the coil end 2 with respect to the predetermined excitation force is sufficiently small. Diagnosed. If the vibration of the coil end 2 is sufficiently small, the vibration intensity of the coil end 2 is acceptable, and the vibration of the coil end 2 is diagnosed as healthy.

ただし、Ｆ_m0・Ｕ_rは、電磁加振力Ｆ_mの振幅Ｆ_m0（ベクトル量）と、コイルエンド２の固有振動モードＵ_r（ベクトル量）との内積を表す。また、係数Ｌ_rは、応答感度（応答倍率）と呼ばれ、次の式（２）で与えられる。

When the vibration frequency f _m of the electromagnetic excitation force F _m approaches the natural frequency f _r of the coil end 2, a large vibration by resonance coil end 2 occurs. Here, representing the natural oscillation mode corresponding to the natural frequency f _r, mode damping ratio, the vibration frequency ratio, respectively U _r, zeta _r, with _{λ r (= f m / f} r). Further, the amplitude of the electromagnetic excitation force F _m is represented by F _m0 . In this case, the mode frequency response φ _r of each coil of the coil end 2 is given by the following equation (1).

Here, F _m0 · U _r represents the inner product of the amplitude F _m0 (vector quantity) of the electromagnetic excitation force F _m and the natural vibration mode U _r (vector quantity) of the coil end 2. The coefficient L _r is called response sensitivity (response magnification) and is given by the following equation (2).

The response sensitivity L _r is an index representing the sensitivity of the coil to the electromagnetic excitation force F _m . If the response sensitivity L _r is low, even if the vibration frequency ratio lambda _r is closer to 1 (i.e., vibration frequency f _m is the coil closer to the natural frequency f _r be a resonance state), Large vibration does not occur in the coil. On the other hand, if the response sensitivity L _r is high, even if the vibration frequency f _m is away from the natural frequency f _r, (referred to as a forced oscillation so) large vibration can occur in the coil.

The response sensitivity L _r of each coil of the coil end 2 is measured by a rotating machine diagnostic apparatus. The impulse hammer 4 has a built-in load cell 3 for measuring the excitation force. The impulse hammer 4 and the acceleration pickup 5 are connected to the FFT analyzer 8 via the power unit 7.

The coil end 2 has first to n-th measurement points in the first to _n-th coils C _{1 to} C _n , respectively. The coil end 2 usually has one measuring point for each coil. The first to n-th measurement points are provided on the same circumference of the coil end 2, and the axial positions (positions in the X direction) of the first to n-th measurement points are all the same.

The measurement of the response sensitivity L _r proceeds from the first coil C ₁ (first measurement point) to the nth coil C _n (nth measurement point). Hereinafter, measurement of the response sensitivity L _r at the k-th coil C _k (k-th measurement point) will be described. k is an integer satisfying 1 ≦ k ≦ n.

ただし、ｆは周波数を表し、Ｆ_k（ｆ）、Ａ_k（ｆ）、Ｈ_k,k（ｆ）はそれぞれ、周波数領域での加振力、応答加速度、伝達関数を表す。このような測定が、第１から第ｎの測定点について順次行われる。なお、第１から第ｎの測定点は、第１から第ｎのコイルＣ₁〜Ｃ_nの半径方向ではなく、第１から第ｎのコイルＣ₁〜Ｃ_nの接線方向や軸方向に設けてもよい。 First, the acceleration pickup 5 is installed at the k-th measurement point located in the radial direction of the k-th coil C _k , and the impact vibration is applied to the k-th measurement point by the impulse hammer 4. Next, a response acceleration A _k (t) with respect to the excitation force F _k (t) given to the kth measurement point is measured by the acceleration pickup 5. Here, t represents time, and F _k (t) and A _k (t) represent excitation force and response acceleration in the time domain, respectively. Next, the excitation force F _k measured by the load cell 3, the response acceleration A _k measured by the accelerometer 5 entered FFT analyzer 8 via the power unit 7 is represented by the following formula (3) The transfer function H _{k, k} (f) is calculated by the FFT analyzer 8.

Here, f represents frequency, and F _k (f), A _k (f), and H _{k, k} (f) represent excitation force, response acceleration, and transfer function in the frequency domain, respectively. Such measurement is sequentially performed for the first to nth measurement points. The measurement points of the first n from the first, not the radial direction of the coil C ₁ -C _n of the n from the first, provided from a first tangentially or axially of the coil C ₁ -C _n of the n May be.

FIG. 9 is a graph for explaining a general rotating machine diagnostic method. In this rotating machine diagnosis method, whether or not the vibration intensity of each coil of the coil end 2 is acceptable is diagnosed based on the response sensitivity L _r . This diagnosis is executed by the arithmetic unit 9.

The response sensitivity L _r of the k-th coil C _k is equivalent to the amplitude of the transfer function H _{k, k} (f). FIG. 9 shows the response sensitivity L _r of the k-th coil C _k measured by the above-described measurement method, using a frequency spectrum display. In FIG. 9, the response sensitivity L _r shows peak values at frequencies f ₁ , f ₂ , f ₃ , and f ₄ .

In the diagnosis method using the response sensitivity L _r , margins Δf _L and Δf _H are taken around the excitation frequency f _m and attention is paid to the frequency range of f _L ≦ f ≦ f _H. Here, f _L and f _H are a lower limit frequency and an upper limit frequency, respectively, and are given by the following equations (4) and (5), respectively.

In the diagnosis method using the response sensitivity L _r, is compared with f _L ≦ f ≦ f judgment value L _cri set in advance the maximum value L _max of the response sensitivity L _r in the frequency range of the _H, the coil of the k evaluate the soundness against vibration frequency f _m of the C _k to (diagnosis). Specifically, if L _max ≦ L _cri, it is determined to be acceptable, and if L _max > L _cri, it is determined to be unacceptable. That is, in FIG. 9, if the response sensitivity L _r is within the region R in the frequency range of f _L ≦ f ≦ f _H , it is determined that the sound is healthy.

The reason for providing a margin Delta] f _L, Delta] f _H is based on the natural frequency f _r of the coil changes with temperature. The vibration of the coil is measured during operation of the rotating machine, and the natural frequency f _r is adjusted by adjusting the temperature and flow rate of the refrigerant that cools the coil so that the response sensitivity L _{r at the} excitation frequency f _m is reduced. May be changed.

FIG. 10 is a diagram illustrating an example of the distribution of electromagnetic excitation force F _m in a general rotating machine.

ここで、ωは回転機械の電源角周波数、ｐは回転機械の極対数、θは回転機械の機械角を表す。 The coil end 2 is a conical shell-like structure projecting outward in the axial direction of the stator 1 and has a number of annular vibration modes as the natural vibration mode _Ur . On the other hand, the electromagnetic excitation force F _m acting on the coil end 2 is generally expressed by the following equation (6).

Here, ω represents the power supply angular frequency of the rotating machine, p represents the number of pole pairs of the rotating machine, and θ represents the mechanical angle of the rotating machine.

FIG. 10A shows the distribution of the electromagnetic excitation force F _m when the rotating machine is a two-pole machine (p = 2). FIG. 10 (a) shows an annular vibration mode having four nodes of vibration on the circumference and a harmonic distribution of two periods along the circumference (annular vibration mode of four nodes and two diameter nodes). Is shown. FIG. 10B shows the distribution of the electromagnetic excitation force F _m when the rotating machine is a quadrupole machine (p = 4). FIG. 10B shows an annular vibration mode of 8 nodes and 4 diameter nodes.

As described above, the mode frequency response φ _r of each coil of the coil end 2 is given by Equation (1). Therefore, the mode frequency response φ _r is given by the product of the response sensitivity L _r and the mode excitation force F _m0 · U _r, and the mode excitation force F _m0 · U _r is the amplitude F of the electromagnetic excitation force F _m . It is given by the inner product of the _m0 and the natural vibration mode U _r.

The electromagnetic excitation force F _m has an annular vibration mode distribution as shown in Equation (6). Moreover, the natural vibration modes U _r of the coil end 2 also has a circular mode of vibration as described above. Since the circular mode of vibration that has a distribution of harmonic functions in the circumferential direction, in the inner product between the amplitude F _m0 and natural vibration modes U _r of the electromagnetic excitation force F _m, it holds the orthogonality of the trigonometric functions.

For example, when the rotating machine is a two-pole machine, the electromagnetic excitation force F _m indicates an annular vibration mode having four nodes and two diameter nodes. Therefore, when the natural vibration mode U _r is an annular vibration mode having four nodes and two diameter nodes, the mode excitation force F _m0 · U _r can have a value other than zero. On the other hand, when the natural vibration mode _Ur is another annular vibration mode, the mode excitation force F _m0 · U _r becomes zero. Thus, natural vibration modes U _r to significantly vibrate the coil 2-pole machine, a circular mode of vibration of the four-node second diameter section, be handled separately from the other natural vibration modes are determined.

Japanese Patent No. 5101213

Shinohara Main Isao "Natural Vibration Characteristics of Turbine Generator Stator Coil End" The Japan Society of Mechanical Engineers, Volume 70, 692, 941-948 (April 2004)

In a general vibration diagnosis of the coil end 2, the soundness of the vibration of the coil end 2 is determined only by the response sensitivity _Lr . However, in order to determine the soundness of the vibration of the coil end 2, it is not sufficient to evaluate the response sensitivity L _r alone, and the natural vibration mode U _r having a mode shape similar to the electromagnetic excitation force F _m is taken into consideration. Is required. To do this, the natural vibration modes U _r of the coil end 2, it is required to extract the circular mode of vibration having the same number of nodes and the number of diameter clause electromagnetic excitation force F _m.

Moreover, because the actual coil end 2 has a complicated shape, the natural oscillation mode U _r is not necessarily become a pure circular mode of vibration, one vibration modes overlap multiple circular mode of vibration It may become. In this case, from this vibration mode, extracts the circular mode of vibration, similar to a circular mode of vibration component of the electromagnetic excitation force F _m, it is required to identify the component ratio of the components.

  Therefore, an object of the present invention is to provide a rotating machine diagnostic apparatus and a rotating machine diagnostic method capable of improving the accuracy of vibration diagnosis of the rotating machine.

  According to one embodiment, the rotating machine diagnostic apparatus has a k-th portion (k is 1 ≦ k ≦ n) of the first to n-th portions (n is an integer of 2 or more) of the structure of the rotating machine. An oscillating unit that applies an oscillating force to the integer) is provided. Furthermore, the apparatus is provided with a first vibration response measuring unit that measures a first vibration response that is a vibration response of the kth part with respect to an excitation force applied to the kth part, and the kth part. A second vibration response measuring unit that measures a second vibration response that is a vibration response of the first portion with respect to the excitation force. Further, the device calculates a first transfer function that is a transfer function of the k-th portion with respect to an excitation force applied to the k-th portion based on the first vibration response, and determines the second vibration response as the second vibration response. And a transfer function calculating unit that calculates a second transfer function that is a transfer function of the first part with respect to the excitation force applied to the kth part. Furthermore, the apparatus includes a diagnosis unit that diagnoses whether the vibration intensity of the structure is acceptable based on the first and second transfer functions.

It is the schematic which shows the structure of the rotary machine diagnostic apparatus of 1st Embodiment. It is a graph which shows the example of the response sensitivity calculated in 1st Embodiment. It is a graph which shows the example of the amplitude and phase of a transfer function computed in a 1st embodiment. It is a graph which shows the example of the natural vibration mode calculated in 1st Embodiment. It is a graph which shows the example of the harmonic coefficient calculated in 1st Embodiment. It is the schematic which shows the structure of the rotary machine diagnostic apparatus of 2nd Embodiment. It is sectional drawing and graph for demonstrating the natural vibration mode measured in 2nd Embodiment. It is the schematic which shows the structure of a general rotary machine diagnostic apparatus. It is a graph for demonstrating the general rotating machine diagnostic method. It is a figure which shows the example of distribution of the electromagnetic exciting force in a common rotary machine.

  Embodiments of the present invention will be described below with reference to the drawings.

(First embodiment)
FIG. 1 is a schematic diagram illustrating the configuration of the rotating machine diagnostic apparatus according to the first embodiment.

  In the configuration of the rotating machine diagnostic apparatus of FIG. 1, the same or similar components as those of the rotating machine diagnostic apparatus of FIG.

FIG. 1A and FIG. 1B show a stator 1 of a rotating machine, a coil end 2 of the stator 1, and a rotating machine diagnostic apparatus that diagnoses the rotating machine. The coil end 2 is an example of a rotating machine structure. The coil end 2 includes first to nth coils C _{1 to} C _n (n is an integer of 2 or more). The first to nth coils C _{1 to} C _n are examples of the first to nth portions of the structure, respectively. An example of the rotating machine is a rotating electric machine such as a turbine generator.

  The rotating machine diagnostic apparatus of the present embodiment includes a load cell 3, an impulse hammer 4 as an example of a vibration unit, an acceleration pickup 5 as an example of a first vibration response measurement unit, and an example of a second vibration response measurement unit. An acceleration pickup 6, a power unit 7, an FFT analyzer 8 that is an example of a transfer function calculation unit, and an arithmetic device 9 that is an example of a diagnosis unit are provided.

  The impulse hammer 4 has a built-in load cell 3 for measuring the excitation force. The impulse hammer 4 and the acceleration pickups 5 and 6 are connected to the FFT analyzer 8 via the power unit 7, and the FFT analyzer 8 is connected to the arithmetic device 9. The arithmetic unit 9 executes various numerical processes using the output information from the FFT analyzer 8. An example of the arithmetic unit 9 is a personal computer.

The coil end 2 of the present embodiment has first to n-th measurement points in the first to _n-th coils C _{1 to} C _n , respectively. Although the coil end 2 of this embodiment has a measurement point in all the coils, you may have a measurement point in only some coils. For example, in order to shorten the work time for vibration diagnosis, only n / n ₀ coils out of n coils may have measurement points (n ₀ is an integer of 2 or more). The first to n-th measurement points of the present embodiment are provided on the same circumference of the coil end 2, and the axial positions (X-direction positions) of the first to n-th measurement points are all the same. It is.

In the vibration diagnosis of this embodiment, the first coil C ₁ (first measurement point) to the nth coil C _n (nth measurement point) are sequentially excited by the impulse hammer 4. Hereinafter, a process for exciting the k-th coil C _k (k-th measurement point) will be described. k is an integer satisfying 1 ≦ k ≦ n.

First, the acceleration pickup 5 is installed at the kth measurement point located in the radial direction of the kth coil _Ck . Further, the acceleration pickup 6 is installed at a first measurement point located in the radial direction of the first coil C ₁ . The first measurement point is a reference point among the first to nth measurement points. In the present embodiment, the coil having the largest Z coordinate is the first coil C ₁ , but other coils may be the first coil C ₁ . That is, the reference point may be provided on a coil other than the coil having the largest Z coordinate.

Next, impulse vibration is performed by the impulse hammer 4 on the kth measurement point. Next, a response acceleration A _k (t) with respect to the excitation force F _k (t) given to the kth measurement point is measured by the acceleration pickup 5. The response acceleration A ₁ (t) with respect to the excitation force F _k (t) given to the kth measurement point is measured by the acceleration pickup 6. Here, t represents time, F _k (t) represents an excitation force in the time domain, and A _k (t) and A ₁ (t) represent response acceleration in the time domain. Response accelerations A _k (t) and A ₁ (t) are examples of first and second vibration responses, respectively.

ただし、ｆは周波数、Ｆ_k（ｆ）は周波数領域での加振力、Ａ_k（ｆ）、Ａ₁（ｔ）は周波数領域での応答加速度、Ｈ_k,k（ｆ）、Ｈ_1,k（ｆ）は周波数領域での伝達関数を表す。伝達関数Ｈ_k,k（ｆ）、Ｈ_1,k（ｆ）はそれぞれ、第１および第２伝達関数の例である。 Next, the excitation force F _k measured by the load cell 3 and the response accelerations A _k (t) and A ₁ (t) measured by the acceleration pickups 5 and 6 are sent to the FFT analyzer 8 via the power unit 7. input. Next, transfer functions H _{k, k} (f) and H _{1, k} (f) represented by the following equations (7) and (8) are calculated by the FFT analyzer 8.

Where f is a frequency, F _k (f) is an excitation force in the frequency domain, A _k (f) and A ₁ (t) are response accelerations in the frequency domain, H _{k, k} (f), H _{1, k} (f) represents a transfer function in the frequency domain. Transfer functions H _{k, k} (f) and H _{1, k} (f) are examples of first and second transfer functions, respectively.

Such measurement is sequentially performed for the first to nth measurement points. At this time, as the excitation position of the impulse hammer 4 changes from the _first coil C ₁ to the n-th coil C _n , the installation position of the acceleration pickup 5 also changes from the first coil C ₁ to the n-th coil. It will change to the coil C _n. On the other hand, even if the excitation position of the impulse hammer 4 is changed from the first coil C ₁ to the n-th coil C _n , the installation position of the acceleration pickup 6 is maintained in the first coil C ₁ .

The measurement points of the first n from the first, not the radial direction of the coil C ₁ -C _n of the n from the first, provided from a first tangentially or axially of the coil C ₁ -C _n of the n May be. However, the installation position of the acceleration pickup 6 of the present embodiment is fixed to the first measurement point located in the radial direction of the first coil C ₁ .

In addition, in the process of measuring the transfer functions H _{k, k} (f) and H _{1, k} (f), it is necessary to perform general frequency analysis using the FFT analyzer 8 such as frequency range setting, resolution setting, window setting, and averaging processing. However, detailed description thereof is omitted.

According to the present embodiment, the transfer function H _1,1 (f), H _2,2 (f),..., H _{n, n} (f) can be calculated by using the acceleration pickup 5. In addition, according to the present embodiment, by using the acceleration pickup 6, the transfer functions H _1,1 (f), H _1,2 (f),..., H _{1, n} (f) can be calculated. .

With respect to the transfer functions H _1,1 (f), H _2,2 (f),..., H _{n, n} (f), H _{k, k} (f) is the excitation force applied to the kth measurement point. The transfer function of the kth measurement point with respect to F _k (t) is shown. The transfer function H _{k, k} (f) _{is, F k (t), A} k (t) respectively F _k (f), is converted to A _k (f), A _k and (f) F _k (f) Calculated by dividing by.

The amplitude of the transfer function H _{k, k} (f) corresponds to the response sensitivity of the k-th coil C _k . Response sensitivity of coil C _k of the k is an index representing a dynamic softness for excitation force of the coil C _k of the k (Accelerated wardrobe). Higher response sensitivity of the coil C _k of the k is large, large vibration in response to the coil C _k of the k occurs for the same excitation force.

FIG. 2 is a graph showing an example of response sensitivity calculated in the first embodiment. Figures. 2 (d) from FIG. 2 (a), the first, second, coil C ₁ of the third and n, C _2, C _3, C response sensitivity of _n (the transfer function H _{1, 1} ( f), H _2,2 (f), H _3,3 (f), H _{n, n} (f) amplitude).

The arithmetic unit 9 of the present embodiment includes first to nth coils C ₁ to C from the transfer functions H _1,1 (f), H _2,2 (f),..., H _{n, n} (f), respectively. Calculate the response sensitivity of _n . And the arithmetic unit 9 of this embodiment diagnoses whether the vibration intensity of each coil is permissible using the response sensitivity of each coil. Specifically, it is diagnosed by the method shown in FIG. 9 whether or not the vibration intensity of each coil is acceptable. Therefore, if the response sensitivity of the k-th coil C _k is within the region R in the frequency range of f _L ≦ f ≦ f _H , the vibration intensity of the k-th coil C _k is sufficient (allowable). It is determined. On the other hand, if the response sensitivity of the k-th coil C _k does not fall within the region R in the frequency range of f _L ≦ f ≦ f _H , the vibration intensity of the k-th coil C _k is insufficient (unacceptable). It is determined that there is.

For the transfer functions H _1,1 (f), H _1,2 (f),..., H _{1, n} (f), H _{1, k} (f) is the excitation force applied to the kth measurement point. The transfer function of the first measurement point with respect to F _k (t) is shown. The transfer function H _{1, k} (f) converts F _k (t) and A ₁ (t) to F _k (f) and A ₁ (f), respectively, and converts A ₁ (f) to F _k (f). Calculated by dividing by.

FIG. 3 is a graph showing an example of the amplitude and phase of the transfer function calculated in the first embodiment. 3 (a) to 3 (d) show the amplitudes of the transfer functions H _1,1 (f), H _1,2 (f), H _1,3 (f), and H _{1, n} (f), respectively. An example of the phase is shown.

The arithmetic unit 9 of this embodiment identifies the natural vibration mode _Ur of the coil end 2 from the transfer functions H _1,1 (f), H _1,2 (f),..., H _{1, n} (f). . The method of identifying a natural vibration mode U _r, transfer of the transfer function H _{1, 1} (f) to H _1, n (f) is the first coil C _k of the coil C ₁ -C _n of the n from the first A simple method for calculating the natural vibration mode U _r by dividing by the function H _1,1 (f) is known. This method is called an operation mode modification method or an ODS method.

ただし、式（９）の導出においては、後述の相反定理Ｈ_k,1（ｆ）＝Ｈ_1,k（ｆ）が使用されている。コイルエンド２の固有振動モードＵ_rは、第１から第ｎのコイルＣ₁〜Ｃ_nの固有振動モード成分ｕ₁〜ｕ_nを用いて、次の式（１０）で表される。

In a production mode variant, the peak value of the transfer function spectral diagram, it is possible to calculate the natural frequency f _r. Furthermore, transmitted to the peak value of the function spectral diagram by applying the half-power method, it is possible to calculate the modal damping factor zeta _r corresponding to the natural frequency f _r. According to a production mode variant, natural vibration mode components u _k of the coil C _k of the k is given by the following equation (9).

However, the reciprocity theorem H _{k, 1} (f) = H _{1, k} (f), which will be described later, is used in the derivation of equation (9). The natural vibration mode U _r of the coil end 2 is expressed by the following equation (10) using the natural vibration mode components u _{1 to} u _n of the first to n-th coils C _{1 to} C _n .

As described above, the arithmetic unit 9 of the present embodiment uses the transfer functions H _1,1 (f), H _1,2 (f),..., H _{1, n} (f) to determine the natural vibration mode U of the coil end 2. Calculate _r . And the arithmetic unit 9 of this embodiment diagnoses whether the whole vibration intensity of the coil end 2 is permissible using the natural vibration mode _Ur of the coil end 2. Details of the method for diagnosing the vibration intensity by the natural vibration mode _Ur will be described later.

According to the present embodiment, the vibration intensity related to the local response of the coil end 2 can be diagnosed based on the response sensitivity of the first to n-th coils C _{1 to} C _n . Further, according to the present embodiment, the vibration intensity related to the overall response of the coil end 2 can be diagnosed based on the natural vibration mode _Ur of the coil end 2. Therefore, according to the present embodiment, it is possible to comprehensively diagnose the diagnostic strength of the coil end 2.

(1) In the present embodiment Experimental Modal Analysis, when calculating the natural vibration mode U _r, it is desirable to perform the experimental modal analysis. In this case, the natural vibration mode _Ur is calculated by curve fitting of the transfer functions H _1,1 (f), H _1,2 (f),..., H _{1, n} (f). As a result, the natural vibration mode _Ur can be accurately calculated. The experiment mode analysis of the present embodiment can be realized, for example, by executing software for experiment mode analysis in the arithmetic unit 9.

In general the experimental modal analysis, natural mode U _r is normalized modal mass m _r, natural oscillation mode U in addition to _r-mode damping factor zeta _r and mode angle natural frequency omega _r is calculated. By dividing the mode angle natural frequency ω _r by 2π, the natural frequency f _r is calculated.

When calculating the natural vibration mode U _r in experimental modal analysis, the transfer function _{H 1,1 (f), H 2,1} (f), ..., common to use H n, ₁ (f) It is. These transfer functions are given repeated excitation force F ₁ to the first coil C ₁ (t), the response acceleration _{A 1 (t) ~A n (} t coil C ₁ -C _n of the first to n ) Can be calculated sequentially. In this case, it is necessary to sequentially move the installation position of the acceleration pickup 5 from the first coil C ₁ to the n-th coil C _n , and the response acceleration measurement work becomes complicated.

However, if the reciprocity theorem is applied, the relationship of the following equation (11) is established between the transfer function H _{k, 1} (f) and the transfer function H _{1, k} (f).

Therefore, in the experimental mode analysis of this embodiment, the natural vibration mode _Ur is calculated using the transfer functions H _1,1 (f), H _1,2 (f),..., H _{1, n} (f). can do. These transfer functions sequentially apply excitation forces F ₁ (t) to F _n (t) to the first to n-th coils C _{1 to} C _n , and the response acceleration A ₁ (t of the first coil C ₁ ) Can be calculated repeatedly.

Figure 4 is a graph showing an example of a natural vibration mode U _r calculated in the first embodiment. According to the experimental modal analysis of the present embodiment, to calculate the natural vibration mode U _r, it is possible to draw such a graph.

FIGS. 4 (a) each of the natural vibration mode U _r of ~ Figure 4 (d), 4 node 2 diameter node mode, 6 nodal third diameter section mode, 8-node 4 diameter clause mode, if it is 10 nodes 5 diameter clause Mode Can be visually recognized.

On the other hand, natural vibration modes U _r shown in FIG. 4 (e) and FIG. 4 (f) it is difficult to recognize the type of mode. In FIG. 4 (e), the excitation point and the response around it are excellent, and the excitation force is not propagated throughout the coil end 2. Therefore, the natural oscillation mode U _r in FIG. 4 (e) is considered to be a local natural vibration modes. Moreover, the natural vibration modes U _r in FIG. 4 (f) appears to 10 node 5 diameter clause mode shows the assertive difficult confusing forms a natural vibration mode U _r and the same mode shown in FIG. 4 (d) .

(2) Harmonic Analysis As described above, when the natural vibration mode _Ur is calculated by the experimental mode analysis, it may be difficult to visually determine the type of the natural vibration mode _Ur . Therefore, it is desirable to quantify the determination of the type of natural vibration modes U _r.

Therefore, the arithmetic unit 9 of the present embodiment performs spatial harmonic analysis of the natural vibration mode _Ur . More specifically, the harmonic analysis of the natural vibration mode _Ur is performed by associating one circumference of the coil end 2 with one period and replacing the space region related to the natural vibration mode _Ur with a time region.

  In harmonic analysis, the time domain is generally converted to the frequency domain. In this case, fast Fourier transform (FFT) is widely used as a harmonic analysis technique. FFT is a method of Fourier transform that is executed at a high speed with a short calculation time. However, the FFT has a restriction that the number of samples (in this embodiment, the number of coils n) is limited to a multiplier of 2 for speeding up.

  On the other hand, since the rotating electrical machine often employs three-phase alternating current, the number of coils of the rotating electrical machine is generally a multiple of six. Therefore, when the harmonic analysis of the rotating electrical machine is performed by FFT, it is necessary to set the number of samples to a multiplier of 2 by interpolation.

However, the number of coils of a general rotating electrical machine is smaller than the general number of samples in the time domain problem that requires high speed. The number of samples in the time domain problem is typically 1024 or more. On the other hand, the number of coils of a rotating electrical machine is generally about 120 at most. Further, the harmonic analysis of the present embodiment, since sufficient if the determination is the mode number of the electromagnetic excitation force mode to excite the natural vibration modes U _r, the order of the required becomes harmonic coefficients harmonic analysis of the present embodiment In many cases, a lower order is sufficient.

ここで、ｒは固有振動モード次数を表し、ｊは調波モード次数を表す。調波モード次数ｊのモードは、振動の節点数が２ｊで直径節数がｊのモード形態を有する。式（１３）と式（１４）に示すように、固有振動モードＵ_rの調波係数は余弦波成分Ａ_jと正弦波成分Ｂ_jとを含むため、本実施形態では、固有振動モードＵ_rの調波係数を、次の式（１５）で与えられる振幅Ｃ_jで代表させる。

ただし、ｊ＝０の場合には、Ｃ₀＝Ａ₀とする。 As described above, the harmonic analysis of this embodiment often does not require high speed. Therefore, the harmonic analysis of this embodiment is performed by discrete Fourier analysis. The discrete Fourier analysis of the present embodiment, the following equation (12), equation (13), harmonic coefficients of the natural oscillation mode U _r in equation (14) is calculated.

Here, r represents the natural vibration mode order, and j represents the harmonic mode order. The mode of the harmonic mode order j has a mode configuration in which the number of vibration nodes is 2j and the diameter node number is j. As shown in the equations (13) and (14), the harmonic coefficient of the natural vibration mode U _r includes a cosine wave component A _j and a sine wave component B _j. Therefore, in this embodiment, the natural vibration mode U _{r Are} represented by the amplitude C _j given by the following equation (15).

However, when j = 0, C ₀ = A ₀ is set.

(3) Harmonic Coefficient FIG. 5 is a graph showing an example of the harmonic coefficient C _j calculated in the first embodiment.

5 (a) to 5 (f) show the harmonic analysis results of the natural vibration modes _Ur of FIGS. 4 (a) to 4 (f), respectively. the harmonic coefficients C _j of the natural oscillation mode harmonic mode constituting a U _r of f) are shown in a bar graph. Calculation device 9 of this embodiment, the harmonic analysis of the natural oscillation mode U _r, calculates the harmonic coefficients C _j natural oscillation modes U _r.

In the graph of FIG. 5A, the secondary harmonic coefficient C ₂ is larger than the other harmonic coefficients C _j . Therefore, it can be seen that the natural vibration mode _Ur of FIG. 5A is a 4-node 2-diameter node mode, as can be visually recognized from FIG. In this manner, when a harmonic coefficients C _j natural oscillation modes U _r, it is possible to quantitatively determine the type of natural vibration modes U _r.

Similarly, in the graph of FIG. 5B, the third-order harmonic coefficient C ₃ is larger than the other harmonic coefficients C _j . In the graph of FIG. 5C, the fourth-order harmonic coefficient C ₄ is larger than the other harmonic coefficients C _j . In the graph of FIG. 5D, the fifth-order harmonic coefficient C ₅ is larger than the other harmonic coefficients C _j . Thus, FIG. 5 (b), the visually recognized from each natural oscillation modes U _r is FIG. 4 (b), the FIG. 4 (c), the Figure 4 (d) of FIG. 5 (c), FIG. 5 (d) It can be seen that the 6-node 3-diameter mode, the 8-node 4-diameter mode, and the 10-node 5-diameter mode are possible.

However, in the graph of FIG. 5A, the first-order harmonic coefficient C ₁ is smaller than the second-order harmonic coefficient C ₂ but larger than the other harmonic coefficients C _j . Thus, natural vibration modes U _r in FIGS. 5 (a) is not a simple four-node second diameter section mode, it can be seen that two nodes first diameter section mode (rigid body modes) is superimposed.

Further, in the graph of FIG. 5 (e), the first-order harmonic coefficient C ₁ and the second-order harmonic coefficient C ₂ have similar values, and these harmonic coefficients C ₁ and C ₂ are the other values. It is larger than the harmonic coefficient C _j . Natural vibration modes U _r in FIG. 5 (e) has been difficult to recognize the type in FIG. 4 (e). However, the natural vibration modes U _r in FIG. 5 (e) from the harmonic analysis result, it can be seen two nodes first diameter section mode and a four-node 2 diameter clause mode is a mode in which a mixture.

In the graph of FIG. 5F, the first, fourth, and fifth harmonic coefficients C ₁ , C ₄ , and C ₅ are larger than the other harmonic coefficients C _j . FIG natural vibration modes U _r of 5 (f) is seen as shown in FIG. 4 (f) in 10 node 5 diameter clause mode. However, the natural vibration modes U _r in FIG. 5 (f) from the harmonic analysis, two nodes first diameter section mode, 8-node 4 diameter clause mode, and 10 node 5 diameter clause mode is a mode in which a mixture I understand. Moreover, the natural vibration modes U _r in FIG. 5 (f), the it can be seen that the proportion of 8-node 4 diameter clause mode is larger than the proportion of 5 nodes 10 diameter clause mode.

(4) As described above diagnosis of vibration intensity, according to the harmonic analysis of the natural oscillation mode U _r, it is quantified by harmonic coefficients C _j the ratio of the harmonic modes which constitute the natural vibration mode U _r it can. Therefore, the arithmetic unit 9 in this embodiment, on the basis of the harmonic coefficients C _j natural oscillation modes U _r calculated by the harmonic analysis, vibration intensity of the coil end 2 to diagnose whether acceptable .

Hereinafter, it is assumed that the rotating machine of this embodiment is a two-pole turbine generator. In this case, the electromagnetic excitation force F _m acting on the coil end 2 is expressed by the following equation (16).

  Equation (16) is derived by substituting p = 2 into Equation (6). The electromagnetic excitation force mode in this case is a mode as shown in FIG. Therefore, the electromagnetic excitation force mode in this case is a 4-node 2-diameter node mode, and has a second-order harmonic component (second-order harmonic coefficient).

The displacement X of the electromagnetic vibration of the coil end 2 is calculated as follows. First, the mode motion equation derived from the experimental mode analysis is solved, and the mode frequency response φ _r is calculated as the solution of the mode motion equation. The mode motion equation is expressed by the following equation (17).

ただし、ｍは、変位Ｘの算出に用いる最大モード次数であり、任意の次数を設定可能である。本実施形態においては、次数ｍを、電磁加振周波数ｆ_mの２倍程度の固有振動数の次数に設定すれば十分である。 The displacement X of the electromagnetic vibration of the coil end 2 is calculated by superimposing the product of the mode frequency response φ _r and the natural vibration mode U _r for a plurality of modes. The displacement X of the electromagnetic vibration of the coil end 2 is expressed by the following formula (18).

However, m is the maximum mode order used for calculating the displacement X, and an arbitrary order can be set. In the present embodiment, it is sufficient to set the order _{m to} the order of the natural frequency that is about twice the electromagnetic excitation frequency fm.

When attention is paid to the right side of the mode motion equation, the magnitude of the excitation force is determined by F _m0 · U _r . When considering the inner product F _m0 · U _r , it is noted that orthogonality is established between the harmonic functions. Since having an electromagnetic excitation force F _m is the 4-node second diameter section mode distribution in this embodiment, it is a second-order harmonic component 4 node 2 diameter node mode component is excited in the coil end 2 as an electromagnetic oscillation, other Harmonic components (eg, 6-node 3-diameter node mode components) are not excited. This is because the contribution of other harmonic components to F _m0 · U _r becomes 0 due to the orthogonality.

Therefore, the displacement X of the electromagnetic vibration of the coil end 2 is dependent on the natural vibration mode U 2 harmonic coefficients C ₂ of _r corresponding to the natural frequency f _r is close to the electromagnetic excitation frequency f _m. For example, the greater the second order harmonic coefficients C ₂ of the natural oscillation mode U _r, large vibration of the coil end 2 4 nodes 2 diameter node mode can occur. On the other hand, the smaller the secondary harmonic coefficients C ₂ of the natural oscillation mode U _r, large vibration of the coil end 2 4 nodes 2 diameter clause mode does not occur.

Thus, the arithmetic unit 9 in this embodiment, among the harmonic coefficients C _j natural oscillation modes U _r calculated by the harmonic analysis, based on the 2-order harmonic coefficients C _2, vibration intensity of the coil end 2 Diagnose whether it is acceptable.

Specifically, the arithmetic unit 9 specifies the f _L ≦ f ≦ f _H natural oscillation modes U _r corresponding to the natural frequency f _r in the frequency range (Fig. 9), of the natural vibration mode U _r A second harmonic coefficient C ₂ is calculated. Next, the arithmetic unit 9, by comparing the second-order harmonic coefficient C ₂ and the preset determination value, to evaluate the soundness to the electromagnetic excitation frequency f _m of the coil end 2 (diagnosis). Specifically, the secondary harmonic coefficient C ₂ is judged to be acceptable if less than or equal to the determination value, it is determined that the vibration intensity of the coil end 2 is sufficient (acceptable). On the other hand, the secondary harmonic coefficient C ₂ is determined to fail is greater than the determination value, it is determined that the vibration intensity of the coil end 2 is insufficient (unacceptable).

Note that the rotating machine diagnosis method of the present embodiment can also be applied to a turbine generator having a pole pair number p other than two. In this case, the arithmetic unit 9 in this embodiment, among the harmonic coefficients C _j natural oscillation modes U _r, based on those having the same order as the harmonic coefficients of the electromagnetic excitation force mode, the vibration of the coil end 2 Diagnose whether the intensity is acceptable. That is, based on the 2p next harmonic coefficients C _2p natural oscillation modes U _r, diagnosis is made.

In addition, the arithmetic device 9 of the present embodiment may output the diagnosis result based on the response sensitivity L _r and the diagnosis result based on the natural vibration mode _Ur , or may output them individually. In the former case, for example, the arithmetic unit 9 outputs that the vibration intensity of the coil end 2 is sufficient when both diagnosis results are acceptable, and the vibration intensity of the coil end 2 is not satisfactory when at least one of the diagnosis results is unacceptable. Outputs enough. In the latter case, for example, the arithmetic unit 9 outputs that the vibration intensity of each coil of the coil end 2 is acceptable, and outputs that the vibration intensity of the entire coil end 2 is unacceptable. In the latter case, the diagnosis results of the first to n-th coils C _{1 to} C _n based on the response sensitivity L _r may be output for each individual coil.

  The rotating machine diagnostic apparatus of the present embodiment includes the acceleration pickups 5 and 6 that measure the vibration acceleration as the vibration response. Instead, the speedometer that measures the vibration speed as the vibration response or the vibration as the vibration response. A displacement meter for measuring the displacement may be provided.

(5) Effects of First Embodiment In this embodiment, the coil that is struck and excited by the impulse hammer 4 and the coil that is measured for vibration by the acceleration pickup 5 are the same. Through this process, the response sensitivity L _{r of} each coil is measured (calculated).

  On the other hand, the acceleration pickup 6 installed at the reference point is fixed to the same coil even if the coil to be struck by the impulse hammer 4 is changed. Through this process, a transfer function representing the correlation between the coil where the reference point is installed and the coil subjected to the impact vibration is measured (calculated).

In the present embodiment, by executing the harmonic analysis of the natural oscillation mode U _r, it is possible to quantify the circular mode of vibration components constituting the natural vibration mode U _r.

According to the present embodiment, by using the two acceleration pickups 5 and 6, it is possible to simultaneously measure the transfer function of the coil at the measurement point subjected to impact vibration and the transfer function of the coil at the reference point. If the hammering excitation of the coil goes around the first to n-th coils C _{1 to} C _n , the natural vibration mode U _r of the coil end 2 is measured together with the response sensitivity L _{r of} each coil.

Further, in practicing the experimental modal analysis in the coil end 2 repeats the oscillation blow pressurized of the first coil C _1, sequentially measuring the response acceleration of the coil C ₁ -C _n of the n from the first Therefore, it is general to calculate the transfer functions H _1,1 (f), H _2,1 (f),..., H _{n, 1} (f). In this case, the impulse hammer 4 strikes only the first coil of the coil C _1. On the other hand, in order to calculate the response sensitivity L _r of each coil, it is necessary to sequentially perform impact excitation of the first to n-th coils C _{1 to} C _n . Therefore, in this case, when the vibration striking pressure for measurement of response sensitivity L _r, it is necessary to perform separately blow vibration for the determination of the natural oscillation mode U _r.

On the other hand, in the experimental mode analysis of the present embodiment, the first to nth coils C _{1 to} C _n are sequentially subjected to striking excitation, and the response acceleration of the first coil C ₁ is repeatedly measured, whereby the transfer function H _1,1 (f), H _1,2 (f),..., H _{1, n} (f) are calculated. Furthermore, in the present embodiment, the first to nth coils C _{1 to} C _n are sequentially subjected to impact excitation, and the response accelerations of the first to nth coils C _{1 to} C _n are sequentially measured, so that The response sensitivity L _r of the 1st to n-th coils C _{1 to} C _n is calculated. Therefore, according to this embodiment, the vibration striking pressure for measurement of response sensitivity L _r, can be shared hitting vibration for the measurement of the natural oscillation mode U _r, response sensitivity L _r and the unique It is possible to measure the vibration mode _Ur in a short measurement operation.

Further, according to this embodiment, by executing the harmonic analysis of the natural oscillation mode U _r, it can be calculated harmonic coefficient C _j of circular mode of vibration components constituting the natural vibration mode U _r. The response sensitivity L _r is an index of the vibration intensity with respect to the electromagnetic vibration for each coil. The harmonic coefficient C _j having the same order as the harmonic coefficient of the electromagnetic excitation force mode is the vibration with respect to the electromagnetic vibration of the entire coil end 2. It is an indicator of strength.

Therefore, according to this embodiment, based on the response sensitivity L _r, local vibration strength of the coil end 2 can diagnose, on the basis of the natural vibration mode U _r, the overall vibration of the coil end 2 Intensity can be diagnosed. Therefore, according to the present embodiment, it is possible to comprehensively diagnose the diagnostic strength of the coil end 2.

As described above, according to the present embodiment, by using the response sensitivity L _r calculated from the transfer function H _{k, k} and the natural vibration mode U _r calculated from the transfer function H _{1, k} , It becomes possible to improve the accuracy of the vibration diagnosis of the rotating machine.

(Second Embodiment)
FIG. 6 is a schematic diagram illustrating the configuration of the rotating machine diagnostic apparatus according to the second embodiment.

  The first to n-th measurement points of the first embodiment are provided on the same circumference of the coil end 2, and all the axial positions (X-direction positions) of the first to n-th measurement points are all provided. The same. In other words, the first to nth measurement points have the same X coordinate. The acceleration pickups 5 and 6 of the first embodiment are installed at the kth measurement point and the first measurement point, respectively (FIG. 1 (a)). Therefore, in FIG. 1A, the position of the acceleration pickup 6 in the X direction is set to the same position as the position of the acceleration pickup 5 in the X direction.

  Similarly to the first embodiment, the first to nth measurement points of the present embodiment are also provided on the same circumference of the coil end 2. However, in this embodiment, the acceleration pickup 5 is installed at the kth measurement point, whereas the acceleration pickup 6 is installed closer to the stator 1 than the first measurement point (FIG. 6A). . Therefore, in FIG. 6A, the position of the acceleration pickup 6 in the X direction is set to a position different from the position of the acceleration pickup 5 in the X direction. The installation positions of the acceleration pickups 5 and 6 in FIG. 6A are examples of first and second locations, respectively.

Figure 7 is a cross-sectional view and a graph for explaining the natural vibration modes U _r measured in the second embodiment.

FIG. 7A shows a cross section of the coil end 2 perpendicular to the Y direction. The natural vibration mode _Ur of the coil end 2 may include an annular vibration mode in which the phase is different between the end portion and the intermediate portion of the coil end 2.

FIG. 7B shows an example of an annular vibration mode in which the phase is the same at the end portion and the intermediate portion. FIG. 7C shows an example of an annular vibration mode in which the phase is different between the end portion and the intermediate portion. Reference sign S _A indicates a deformation of the coil end 2 in the end section AA. Code S _B shows a variation of the coil end 2 in a section B-B of the intermediate portion. The annular vibration mode in FIGS. 7B and 7C is a 4-node 2-diameter node mode. The annular vibration mode of FIG. 7B has the same deformation direction in the cross-section AA and the cross-section BB, and is called an in-phase mode. The annular vibration mode in FIG. 7C is called a reverse phase mode because the direction of deformation is shifted by 90 degrees between the section AA and the section BB.

On the other hand, in the electromagnetic excitation force mode, the phase angle generally changes slightly along the curvature of the coil end 2, but can be regarded as an almost in-phase mode. Accordingly, circular mode of vibration (reverse phase mode) of FIG. 7 (c), hard to be excited by the electromagnetic exciting force as a natural mode U _r. Therefore, when performing vibration diagnostics using harmonic coefficients C _j natural oscillation modes U _r, radial vibration mode of Fig. 7 (c) it is desirable to exclude from use.

In the present embodiment, the acceleration pickup 5 is installed on the section AA, and the acceleration pickup 6 is installed on the section BB. Therefore, the arithmetic unit 9 of the present embodiment is based on the transfer function H _{k, k} (f) and the transfer function H _{1, k} (f), and the circular vibration mode (natural vibration mode U _r ) of the coil end 2. Can be discriminated whether in the common mode or the reverse phase.

Therefore, according to this embodiment, when the vibration diagnostics using harmonic coefficients C _j natural oscillation modes U _r, it is possible to exclude harmonic coefficients C _j of the reverse phase mode from use The vibration diagnosis can be performed using only the harmonic coefficient C _j of the common mode. Therefore, according to this embodiment, with the exclusion of that is never excited natural oscillation modes U _r it is possible to perform vibration diagnostics, it is possible to improve the diagnostic accuracy.

  Although several embodiments have been described above, these embodiments are presented as examples only and are not intended to limit the scope of the invention. The novel apparatus and methods described herein can be implemented in a variety of other forms. In addition, various omissions, substitutions, and changes can be made to the forms of the apparatus and method described in the present specification without departing from the spirit of the invention. The appended claims and their equivalents are intended to include such forms and modifications as fall within the scope and spirit of the invention.

1: Stator, 2: Coil end, 3: Load cell,
4: Impulse hammer, 5, 6: Accelerometer,
7: Power unit, 8: FFT analyzer, 9: Arithmetic unit

Claims (8)

An excitation unit that applies an excitation force to the k-th portion (k is an integer satisfying 1 ≦ k ≦ n) of the first to n-th portions (n is an integer of 2 or more) of the structure of the rotating machine;
A first vibration response measuring unit that measures a first vibration response that is a vibration response of the k-th part with respect to an excitation force applied to the k-th part;
A second vibration response measuring unit that measures a second vibration response that is a vibration response of the first part with respect to an excitation force applied to the kth part;
Based on the first vibration response, a first transfer function, which is a transfer function of the k-th portion with respect to the excitation force applied to the k-th portion, is calculated, and based on the second vibration response, the k-th portion A transfer function calculating unit that calculates a second transfer function that is a transfer function of the first part with respect to the excitation force applied to the part;
A diagnostic unit for diagnosing whether or not the vibration intensity of the structure is acceptable based on the first and second transfer functions;
Rotating machine diagnostic device.   The diagnostic unit calculates a response sensitivity of the kth portion based on the first transfer function, and whether or not the vibration intensity of the kth portion of the structure is acceptable based on the response sensitivity. The rotating machine diagnosis apparatus according to claim 1, wherein the diagnosis is performed.   The diagnostic unit calculates a natural vibration mode of the structure based on the second transfer function, and determines whether the overall vibration intensity of the structure is acceptable based on the natural vibration mode. The rotating machine diagnostic apparatus according to claim 1, which diagnoses.   The rotating machine diagnosis apparatus according to claim 3, wherein the diagnosis unit performs harmonic analysis of the natural vibration mode.   The rotation according to claim 4, wherein the diagnosis unit diagnoses whether the vibration intensity of the structure is acceptable based on a harmonic coefficient of the natural vibration mode calculated by the harmonic analysis. Machine diagnostic device.   The diagnostic unit can allow the vibration intensity of the structure based on the harmonic coefficient of the natural vibration mode having the same order as the harmonic coefficient of the electromagnetic excitation force mode of the rotating machine. The rotating machine diagnostic apparatus according to claim 5, which diagnoses whether or not. The first vibration response measurement unit is installed at a first location in the k-th portion when measuring the first vibration response,
The second vibration response measurement unit is installed at a second location in the first portion when measuring the second vibration response,
The rotary machine diagnostic apparatus according to claim 1, wherein the position of the second location is set to a position different from the location of the first location with respect to the axial position of the rotary machine. . An excitation force is applied to the k-th portion (k is an integer satisfying 1 ≦ k ≦ n) among the first to n-th portions (n is an integer of 2 or more) of the structure of the rotating machine,
Measuring a first vibration response which is a vibration response of the k-th part with respect to an excitation force applied to the k-th part;
Measuring a second vibration response which is a vibration response of the first portion with respect to an excitation force applied to the k-th portion;
Based on the first vibration response, a first transfer function, which is a transfer function of the k-th portion with respect to the excitation force applied to the k-th portion, is calculated, and based on the second vibration response, the k-th portion Calculating a second transfer function which is a transfer function of the first part with respect to the excitation force applied to the part;
Diagnosing whether the vibration intensity of the structure is acceptable based on the first and second transfer functions;
Rotating machine diagnostic method.

Images