CN109271684B

CN109271684B - Dynamic characteristic analysis method for water pump turbine in load shedding process

Info

Publication number: CN109271684B
Application number: CN201810995376.9A
Authority: CN
Inventors: 刘全忠; 苏文涛; 夏煜星; 王兴茹; 徐科繁
Original assignee: Harbin Institute of Technology
Current assignee: Harbin Institute of Technology
Priority date: 2018-08-29
Filing date: 2018-08-29
Publication date: 2022-04-05
Anticipated expiration: 2038-08-29
Also published as: CN109271684A

Abstract

The invention provides a dynamic characteristic analysis method for a pump turbine in a load shedding process, and belongs to the technical field of pump turbine characteristic analysis. The analysis method firstly sets and calculates boundary conditions for unsteady flow field calculation in the load shedding transition process according to a water pump turbine set rotor dynamic equation, then carries out detailed time-frequency characteristic analysis on external characteristics of the water pump turbine set and pressure pulsation of different component measuring points in the whole flow field to obtain four special frequency components a, b, c and d causing unstable external characteristics of the water pump turbine, preliminarily analyzes the excitation source positions of the four special frequency components, and then gives out inducements of the special frequency components from the perspective of the external characteristics. And finally, an internal flow field in the load shedding transition process is analyzed, and specific reasons for forming special frequency components are given from two aspects of vorticity and flow state of a draft tube.

Description

Dynamic characteristic analysis method for water pump turbine in load shedding process

Technical Field

The invention relates to a dynamic characteristic analysis method for a pump turbine in a load shedding process, and belongs to the technical field of pump turbine characteristic analysis.

Background

The most important component of the pumped storage power station is a pump turbine, and the pump turbine and the turbine have double roles of the pump and the turbine, so that the pump turbine can undergo a complex transition process in the operation process, the research and the start of the pump turbine transition process in China are late, and the characteristic of the transition process is one of key factors for measuring the stability of the pumped storage power station, so that the deep research on the pump turbine transition process has important engineering practical significance for the optimal design, safe and stable operation and full play of the economic benefit of the pump turbine.

The existing characteristic analysis method for the operation process of the water pump turbine basically selects a series of working condition points to perform steady state calculation to cover four quadrants, and on the other hand, the characteristic analysis method mainly focuses on a water turbine reverse S area and a water pump hump area. But the research and analysis on frequent switching among working conditions, pressure pulsation characteristics and evolution rules of various special flows in the continuous process of the pump turbine transition process are lacked. In the research on the transient process of the water turbine, much attention is paid to how to accurately realize the transient transition process of the water turbine and research external characteristics in the transient transition process of the water turbine by using CFD numerical simulation, and aiming at the fact that the external characteristics in the transient transition process of the water turbine can be changed violently, and the causes generated by the changes are rarely analyzed in a relevant way.

Disclosure of Invention

In order to solve the defects in the prior art, the invention provides a dynamic characteristic analysis method for a pump turbine in a load shedding process, which adopts the following technical scheme:

a dynamic characteristic analysis method for a pump turbine in a load shedding process comprises the following steps:

the method comprises the following steps: establishing a three-dimensional transient calculation model of the water pump turbine by respectively taking a volute, a fixed guide vane, a movable guide vane, a rotating wheel and a draft tube as calculation areas, dispersing the calculation areas of the three-dimensional transient calculation model by using a grid division method, and determining the number of grid nodes through grid independence verification;

step two: boundary conditions, calculation settings and turbulence models of five calculation areas, namely a volute, a fixed guide vane, a movable guide vane, a rotating wheel and a draft tube in the three-dimensional transient calculation model are respectively set;

step three: calculating the rotating speed of a rotating wheel according to rotor dynamics, and calculating a load shedding transition process on the basis of the rotating speed of the rotating wheel;

step four: arranging pressure measuring points in the full flow field of the pump-turbine, obtaining the pressure pulsation characteristic of the pump-turbine in the load shedding process through the time domain characteristic of the pressure measuring points, analyzing the dynamic characteristic parameters of the pump-turbine in the load shedding process by using the obtained pressure pulsation characteristic, determining that a pressure signal of the pressure measuring points contains a special frequency component a, a special frequency component b, a special frequency component or a special frequency component d, and preliminarily determining an excitation position source of each component contained in the pressure signal;

step five: determining inducement of the special frequency component a, the special frequency component b, the special frequency component c and the special frequency component d according to dynamic characteristic parameters of the water pump turbine in the load shedding process, flow and runner moment external characteristics of the water pump turbine in the load shedding process and comprehensive external characteristics of the load shedding process;

step six: the method comprises the steps of analyzing the vortex quantity evolution process and the flow state in a tail water pipe in the load shedding process of the water pump turbine by utilizing the vortex quantity distribution of the water pump turbine and the pressure signal spectrogram of the water pump turbine in the load shedding process of the water pump turbine, and obtaining specific reasons generated by a special frequency component a, a special frequency component b, a special frequency component c and a special frequency component d from the two aspects of the vortex quantity evolution process and the flow state in the tail water pipe, so as to complete the dynamic characteristic analysis of the water pump turbine in the load shedding process.

Further, the process of setting the boundary conditions, calculation setting and turbulence model in the second step includes:

the first step is as follows: boundary conditions of five calculation areas of a volute, a fixed guide vane, a movable guide vane, a rotating wheel and a draft tube in the three-dimensional transient calculation model are set respectively, and the boundary conditions are specifically set as follows: setting the boundary condition of the volute inlet section as Pressure-inlet; setting the boundary condition of the outlet section of the draft tube as Pressure-outlet; a runner area adopts a slippage grid model; the volute outlet, the fixed guide vane inlet, the movable guide vane outlet, the movable guide vane inlet, the movable guide vane outlet, the rotating wheel inlet, the rotating wheel outlet and the draft tube inlet are set to be 4 pairs of interface boundary conditions; calculating a steady-state initial field by using the parameters of the initial working condition point as an initialization condition of subsequent transient calculation;

the second step is that: the turbulence model is set as follows: the turbulent flow model adopts an RNGk-epsilon model and a SIMPLEC algorithm to carry out numerical solution on a flow control equation;

the third step: performing calculation setting, specifically setting as follows: the convergence residual errors of all parameters in the numerical calculation are set to be 1.0 e-5; in the unsteady state calculation, the time step size is 0.0017 seconds, and the maximum iteration step number of each time step is set to 30 steps.

Further, the step three of calculating the rotating speed of the rotating wheel comprises the following steps:

obtaining a rotor angular momentum balance equation according to rotor dynamics, wherein the rotor angular momentum balance equation is as follows:

wherein M is the resultant moment (Nm) to which the rotor is subjected;

j-rotor moment of inertia (kgm)²)；

ω -rotor angular velocity (rad/s);

t-time(s);

and (3) dispersing the rotor angular momentum balance equation, compiling the dispersed equation form into a Fluent user-defined function by using C language, loading and compiling in Fluent to control the rotation of the rotating wheel, calling in each time step iteration process, and outputting the rotating speed and the torque parameter of each step.

Further, the arrangement positions of the pressure measuring points in the fourth step comprise: in a plane passing through the middle height of the double-row cascade, volute measuring points SC-MP-1-4, fixed guide vane inlet measuring points sv-1-6, movable guide vane inlet measuring points gv-1-20, bladeless area measuring points vl-1-20 and rotating wheel inlet measuring points rn-1-9 are respectively arranged along the clockwise direction: in a plane passing through the movable guide vane and close to the height of the bottom ring, fixed guide vane inlet measuring points sv-dn-1-6, movable guide vane inlet measuring points gv-dn-1-20, bladeless area measuring points vl-dn-1-20 and runner inlet measuring points rn-dn-1-9 are respectively arranged along the clockwise direction: the height of the transitional guide vane close to the top cover is respectively provided with fixed guide vane inlet measuring points sv-up-1-6, movable guide vane inlet measuring points gv-up-1-20, bladeless area measuring points vl-up-1-20 and rotating wheel inlet measuring points rn-up-1-9 in the clockwise direction: monitoring points DT 1-12 are respectively arranged on three different sections of the draft tube.

Further, the step four of obtaining the pressure pulsation characteristic of the pump turbine in the load shedding process comprises the following specific steps:

step 1: obtaining a time domain diagram of a pressure signal of a monitoring point vl1 of a bladeless area through three-dimensional CFD calculation, carrying out fast Fourier transform on the pressure signal of the monitoring point vl1 of the bladeless area to obtain a fast Fourier transform result corresponding to the pressure signal of the monitoring point vl1 of the bladeless area, and determining components contained in the pressure signal of the monitoring point vl1 of the bladeless area according to the fast Fourier transform result; carrying out short-time Fourier transform on the pressure signal of the monitoring point vl1 of the bladeless area, and preliminarily determining an excitation position source of components contained in the pressure signal according to a short-time Fourier transform result corresponding to the pressure signal of the monitoring point vl1 of the bladeless area; the components contained in the pressure signal of the monitoring point vl1 of the bladeless area comprise a special frequency component a and a special frequency component b;

step 2: obtaining a time domain diagram of a pressure signal of a movable guide vane inlet monitoring point gv1 through three-dimensional CFD calculation, carrying out fast Fourier transform and short-time Fourier transform on the pressure signal of the movable guide vane inlet monitoring point gv1, and determining components contained in the pressure signal of the movable guide vane inlet monitoring point gv1 and an excitation position source of the components according to the fast Fourier transform and short-time Fourier transform results; wherein, the components contained in the pressure signal of the guide vane inlet monitoring point gv1 include a special frequency component a, a special frequency component b, a special frequency component c and a special frequency component d;

and step 3: obtaining a time domain diagram of a pressure signal of a fixed guide vane inlet monitoring point sv1 through three-dimensional CFD calculation, carrying out fast Fourier transform and short-time Fourier transform on the pressure signal of the fixed guide vane inlet monitoring point sv1, and determining components contained in the pressure signal of the fixed guide vane inlet monitoring point sv1 and an excitation position source of the components according to the results of the fast Fourier transform and the short-time Fourier transform; the components contained in the pressure signal of the fixed guide vane inlet monitoring point sv1 comprise a special frequency component a, a special frequency component b, a special frequency component c and a special frequency component d;

and 4, step 4: obtaining a time domain diagram of a pressure signal of a volute monitoring point sp1 through three-dimensional CFD calculation, performing fast Fourier transform and short-time Fourier transform on the pressure signal of the volute monitoring point sp1, and determining components contained in the pressure signal of the volute monitoring point sp1 and an excitation position source of the components according to the results of the fast Fourier transform and the short-time Fourier transform; the components contained in the pressure signal of the fixed guide vane inlet monitoring point sv1 comprise a special frequency component a, a special frequency component b, a special frequency component c and a special frequency component d;

and 5: obtaining a time domain diagram of a pressure signal of a monitoring point rn1 of a rotating wheel inlet through three-dimensional CFD calculation, carrying out fast Fourier transform and short-time Fourier transform on the pressure signal of the monitoring point rn1 of the rotating wheel inlet, and determining components contained in the pressure signal of the monitoring point rn1 of the rotating wheel inlet and an excitation position source of the components according to the results of the fast Fourier transform and the short-time Fourier transform; the components contained in the pressure signal of the monitoring point rn1 at the entrance of the rotating wheel comprise a special frequency component a and a special frequency component b;

step 6: obtaining a time domain diagram of a pressure signal of a draft tube inlet measuring point dt1 through three-dimensional CFD calculation, performing fast Fourier transform and short-time Fourier transform on the pressure signal of the draft tube inlet measuring point dt1, and determining components contained in the pressure signal of the draft tube inlet measuring point dt1 and an excitation position source of the components according to the results of the fast Fourier transform and the short-time Fourier transform; the pressure signal of the draft tube inlet measuring point dt1 contains components including a special frequency component a, a special frequency component b, a special frequency component c and a special frequency component d;

and 7: obtaining a time domain diagram of a pressure signal of a draft tube straight-conical section measuring point dt5 through three-dimensional CFD calculation, carrying out fast Fourier transform and short-time Fourier transform on the pressure signal of the draft tube straight-conical section measuring point dt5, and determining components contained in the pressure signal of the draft tube straight-conical section measuring point dt5 and an excitation position source of the components according to the fast Fourier transform and short-time Fourier transform results; the pressure signal of the straight-conical section measuring point dt5 of the draft tube comprises components including a special frequency component a, a special frequency component b, a special frequency component c and a special frequency component d;

and 8: obtaining a time domain diagram of a pressure signal of a measuring point dt9 of the elbow section of the draft tube through three-dimensional CFD (computational fluid dynamics) calculation, carrying out fast Fourier transform and short-time Fourier transform on the pressure signal of the measuring point dt9 of the elbow section of the draft tube, and determining components contained in the pressure signal of the measuring point dt9 of the elbow section of the draft tube and an excitation position source of the components according to the results of the fast Fourier transform and the short-time Fourier transform; the pressure signal of the elbow section measuring point dt9 of the draft tube comprises a special frequency component a, a special frequency component b, a special frequency component c and a special frequency component d.

The invention has the beneficial effects that:

compared with the traditional analysis method of the pump turbine, the analysis method of the dynamic characteristics of the pump turbine in the load shedding process provided by the invention has the advantages that the three-dimensional flow field calculation of the pump turbine is combined with the machine set rotor dynamic equation, the dynamic characteristics of the turbine in the speed fluctuation process under the load shedding working condition of the machine set and the internal flow characteristics of a flow channel of the pump turbine are effectively analyzed, the time-frequency characteristics of pressure pulsation at each measuring point in the water pump turbine under the condition of load shedding speed fluctuation can be obtained by arranging the pressure pulsation measuring point in the flow channel, the pressure pulsation frequency component causing unstable external characteristics in the load shedding process is found, and the internal relation between the special pressure pulsation frequency component and the flow field structure is established through the vorticity and the streamline distribution condition on the intermediate flow surface of the rotating wheel area at different moments.

Drawings

FIG. 1 is a flow chart of a unit runner speed calculation algorithm.

FIG. 2 is a schematic diagram of the distribution of pressure monitoring points in the draft tube.

FIG. 3 is a schematic diagram of distribution of pressure monitoring points of the movable guide vane middle height plane.

Fig. 4 is a schematic distribution diagram of three different height plane pressure monitoring points of the movable guide vane.

FIG. 5 is a graph of the FFT spectrum of the pressure signal at point vl 1.

FIG. 6 is a graph of the STFT spectrum of the pressure signal at point vl 1.

FIG. 7 is a graph of the FFT spectrum of the pressure signal at point gv 1.

FIG. 8 is a graph of the STFT spectrum of the pressure signal at measurement point gv 1.

FIG. 9 is a graph of the FFT spectrum of the pressure signal at point sv 1.

FIG. 10 is a graph of the STFT spectrum of the pressure signal at point sv 1.

FIG. 11 is a graph of the FFT spectrum of the pressure signal at measurement point sc 1.

FIG. 12 is a graph of the STFT spectrum of the pressure signal at measurement site sc 1.

FIG. 13 is a graph of the FFT spectrum of the pressure signal at the measurement point rn 1.

FIG. 14 is a graph of the STFT spectrum of the pressure signal at point rn 1.

FIG. 15 is a graph of the FFT spectrum of a pressure signal at point dt 1.

FIG. 16 is a graph of the STFT spectrum of the pressure signal at point dt 1.

FIG. 17 is a graph of the FFT spectrum of a pressure signal at point dt 5.

FIG. 18 is a graph of the STFT spectrum of the pressure signal at point dt 5.

FIG. 19 is a graph of an FFT spectrum of a pressure signal at point dt 9.

FIG. 20 is a graph of the STFT spectrum of the pressure signal at point dt 9.

Fig. 21 is a flow rate variation process diagram of the water turbine.

FIG. 22 shows a process of torque variation of the rotor.

FIG. 23 is Q₁₁-n₁₁And (5) integrating characteristic graphs.

FIG. 24 is T₁₁-n₁₁And (5) integrating characteristic graphs.

Fig. 25 is a vorticity distribution cloud chart of the mid-height flow surfaces of the fixed guide vane, the movable guide vane and the runner at the time t being 0.20 seconds at each selected time.

Fig. 26 is a vorticity distribution cloud chart of the fixed guide vane, the movable guide vane and the runner middle height flow surface at each selected moment t ═ 3.82 seconds.

Fig. 27 is a vorticity distribution cloud chart of the mid-height flow surfaces of the fixed guide vane, the movable guide vane and the runner at the time t being 5.10 seconds at each selected time.

Fig. 28 is a vorticity distribution cloud plot of the mid-height flow surfaces of the fixed guide vanes, the movable guide vanes and the runner at each selected time instant t ═ 6.44 seconds.

Fig. 29 is a vorticity distribution cloud chart of the mid-height flow surfaces of the fixed guide vane, the movable guide vane and the runner at the moment t ═ 7.00 seconds at each selected moment.

Fig. 30 is a vorticity distribution cloud chart of the mid-height flow surfaces of the fixed guide vane, the movable guide vane and the runner at the moment t ═ 9.40 seconds at each selected moment.

Fig. 31 is a draft tube streamline distribution diagram at each time in the load shedding transition process at the time when t is 0.02 seconds.

FIG. 32 is a draft tube streamline distribution diagram at each time in the load shedding transition process at the time t of 3.94 seconds.

Fig. 33 is a draft tube streamline distribution diagram at each moment in the load shedding transition process at the moment t-6.02 seconds.

FIG. 34 is a draft tube streamline distribution diagram at each moment in the load shedding transition process at the moment t of 7.48 seconds.

Detailed Description

The present invention will be further described with reference to the following specific examples, but the present invention is not limited to these examples.

Example 1:

And step two, the setting processes of the boundary conditions, the calculation setting and the turbulence model comprise:

The calculation algorithm of the rotating speed of the rotating wheel in the third step is shown in fig. 1, and the calculation process comprises the following steps:

wherein M is the resultant moment (Nm) to which the rotor is subjected;

j-rotor moment of inertia (kgm)²)；

ω -rotor angular velocity (rad/s);

t-time(s);

and (3) dispersing the rotor angular momentum balance equation, compiling the dispersed equation form into a Fluent user-defined function by using C language, loading and compiling in Fluent to control the rotation of the rotating wheel, calling in each time step iteration process, and outputting the rotating speed and the torque parameter of each step. The most important judgment basis for the accuracy of the calculation of the load shedding transition process is the simulated rotating speed of the unit, and the error between the rotating speed of the rotating wheel simulated by the rotating speed calculation method of the rotating wheel and the test value is always kept within 5%, so that the accuracy of the rotating speed simulation of the rotating wheel is improved, and the calculation accuracy of the load shedding transition process is further improved.

Step four, the arrangement positions of the pressure measuring points comprise: in a plane passing through the middle height of the double-row cascade, volute measuring points SC-MP-1-4, fixed guide vane inlet measuring points sv-1-6, movable guide vane inlet measuring points gv-1-20, bladeless area measuring points vl-1-20 and rotating wheel inlet measuring points rn-1-9 are respectively arranged along the clockwise direction: in a plane passing through the movable guide vane and close to the height of the bottom ring, fixed guide vane inlet measuring points sv-dn-1-6, movable guide vane inlet measuring points gv-dn-1-20, bladeless area measuring points vl-dn-1-20 and runner inlet measuring points rn-dn-1-9 are respectively arranged along the clockwise direction: the height of the transitional guide vane close to the top cover is respectively provided with fixed guide vane inlet measuring points sv-up-1-6, movable guide vane inlet measuring points gv-up-1-20, bladeless area measuring points vl-up-1-20 and rotating wheel inlet measuring points rn-up-1-9 in the clockwise direction: monitoring points DT 1-12 are respectively arranged on three different sections of the draft tube.

Step four, the concrete steps of obtaining the pressure pulsation characteristic of the pump turbine in the load shedding process are as follows:

The time-frequency analysis of pressure signals of a bladeless area monitoring point vl1, a movable guide vane inlet monitoring point gv1, a fixed guide vane inlet monitoring point sv1, a volute monitoring point sp1, a rotating wheel inlet monitoring point rn1, a draft tube inlet measuring point dt1, a draft tube straight cone measuring point dt5 and a draft tube elbow section measuring point dt9 is specifically as follows:

pressure signal time-frequency analysis of monitoring points of the bladeless area: calculating a time domain diagram of the obtained pressure signal of the monitoring point vl1 of the bladeless area, and performing Fast Fourier Transform (FFT) on the pressure signal in order to determine the frequency components of the pressure signal, wherein the fast Fourier transform result is shown in FIG. 5, and can be determined according to the fast Fourier transform result: the pressure signals of the monitoring points of the bladeless area comprise three high-frequency bands with high amplitude of 56-73 Hz, 113-148 Hz and 182-216 Hz and a series of high-amplitude low-frequency components with high amplitude of 0-50 Hz. The frequency conversion is 6.25-8.125 Hz (f) in the load shedding process₀～1.3f₀) It can be determined that the three higher amplitude high frequency bands correspond approximately to 9, 18 and 27 revolutions, respectively, thereby indicating that the three higher amplitude high frequency bands are caused by dynamic and static interference between the rotor and the stationary component. In order to determine the dynamic and static interference effect and analyze the source of a series of high-amplitude and low-frequency components with the frequency of 0-50 Hz, the pressure signal at the vl1 measuring point is subjected to short-time Fourier transform (STFT), and the result is shown in FIG. 5. As can be seen from the figure, the pressure signal has obvious dynamic and static interference effects, which are expressed as three frequency bands of 9 times, 18 times and 27 times of frequency conversion, and correspond to the result of FFT analysis. There are also three frequency bands opposite to the 9, 18 and 27 times trans-frequency waveforms. In addition, in a period of about 4-7 seconds, high-amplitude frequency components a and b corresponding to a series of high-amplitude low-frequency components of 0-50 Hz in the FFT spectrogram (figure 5) exist in the STFT spectrogram (figure 6), and the water torque of the runner is zero and reaches the vicinity of the maximum runaway rotating speed. Therefore, the influence of the external characteristics in the time period of 4-7 seconds is determined, and the source of the series of high-amplitude continuous low-frequency components with the amplitude of 0-50 Hz caused by the cut-in analysis of the flow field in the time period is determined.

Pressure signal time-frequency analysis of movable guide vane inlet monitoring points: and calculating a time domain diagram of the pressure signal of the movable guide vane inlet monitoring point gv1, and performing Fast Fourier Transform (FFT) and Short Time Fourier Transform (STFT) on the pressure signal of the measuring point gv1 by referring to the time-frequency analysis process of the pressure signal of the monitoring point vl1 of the bladeless area in order to achieve the purpose of performing time-frequency analysis on the pressure signal of the movable guide vane inlet monitoring point gv 1. The Fast Fourier Transform (FFT) and short-time fourier transform (STFT) results are shown in fig. 7 and 8, and the FFT results can determine that, in the pressure signal at the movable guide vane inlet gv1, only 9 frequency doubling components (56-73 Hz) have higher pulsation amplitude in terms of integral frequency doubling components of the rotating frequency, but the pulsation amplitude is much lower relative to the pulsation amplitude of a bladeless region, and it can be determined that the closer to the rotating wheel, the larger the influence, and vice versa, the pressure pulsation components of integral frequency doubling are caused by dynamic and static interference of the rotating wheel. Similarly, a series of high-amplitude continuous low-frequency components with the amplitude of 0-50 Hz similar to the bladeless zone measuring point vl1 also exist in the FFT result, but the maximum amplitude of the pulsation is higher than that of the bladeless zone measuring point vl1, so that the series of high-amplitude low-frequency components with the amplitude of 0-50 Hz can be determined not to be caused by dynamic and static interference. From the STFT results shown in fig. 8, it can be determined that the STFT results show that the spectral distribution about the runner integer multiples is substantially consistent with the FFT analysis results, but there are no frequency bands opposite to the 9-, 18-, and 27-times runner waveforms in the STFT spectrogram of gv1, which further illustrates the strong influence of the dynamic and static interference effects on the bladeless region. Meanwhile, the special low-frequency components of a series of high-amplitude pulsation within the range of about 0-50 Hz within the time period of 4-7 seconds are not as strong as those of the bladeless region, so that the excitation source of the special low-frequency components of the series of high-amplitude pulsation within the range of 0-50 Hz can be determined to be in the bladeless region. In addition, a group of continuous high-frequency high-amplitude pulsation frequency components c of about 50-250 Hz and a group of continuous high-frequency high-amplitude pulsation frequency components d of about 100-250 Hz are also present in the pressure signal of the movable guide vane inlet measuring point gv1 in a period of about 7-7.5 seconds, and the two groups of frequency components do not exist in the bladeless area, so that the excitation source of the movable guide vane is inferred not to exist in the bladeless area.

And (3) pressure signal time-frequency analysis of a fixed guide vane inlet monitoring point: and calculating a time domain diagram of the pressure signal of the fixed guide vane inlet monitoring point sv1, and performing Fast Fourier Transform (FFT) and Short Time Fourier Transform (STFT) on the pressure signal of the measuring point sv1 by analogy with the same purpose and method as the movable guide vane region analysis. From the FFT results, as shown in fig. 9, it can be seen that there is almost no blade passing frequency and its harmonic frequency caused by the dynamic and static interference of the runner in the pressure signal at the inlet of the stay vane and at the measurement point sv1, because the stay vane region is far enough away from the runner, so that the dynamic and static interference effect of the runner hardly propagates. However, a series of continuous low-frequency components with high amplitude of 0-50 Hz similar to the non-leaf region measuring point vl1 and the movable guide vane region measuring point gv1 still exist, the maximum pulsation amplitude is higher than the measuring points vl1 and gv1, and the gradual rising change relation of the maximum pulsation amplitude of the continuous low frequency of 0-50 Hz of the pressure pulsation of the three measuring points vl1, gv1 and sv1 can be preliminarily deduced, so that the existence position of the excitation source with the continuous low frequency of 0-50 Hz can be obtained. From the STFT results (fig. 10), it can be concluded that the 18-fold harmonic due to the dynamic-static interference in the stationary vane region farther from the bladeless region is stronger than the 9-fold harmonic. The continuous low-frequency components a and b with higher amplitude of about 5-50 Hz existing between 4-7 seconds are gradually reduced compared with the points measured by vl1 and gv1, which further indicates that the vibration source of the continuous low frequency of 5-50 Hz is closer to the bladeless area. And the continuous high-frequency component c with higher amplitude of about 50-250 Hz existing between 7-7.5 seconds is stronger than that of the fixed guide vane region. The frequency components gradually increase from zero to a little from a point of vl1 to sv1, which shows that the vibration source of continuous high-frequency components with higher amplitude of about 50-250 Hz exists between 7-7.5 seconds and is closer to the fixed guide vane area. The same stationary guide vane region also has a high-frequency high-amplitude frequency component d.

Pressure signal time-frequency analysis of volute monitoring points: calculating a time domain diagram of the obtained pressure signal of a volute monitoring point sp1, and performing Fast Fourier Transform (FFT) and Short Time Fourier Transform (STFT) on the pressure signal of a measuring point sp 1; the FFT and STFT results are shown in fig. 11 and 12, respectively. The pressure signal condition of a measuring point sp1 can be determined through FFT and STFT results to be basically consistent with the analyzed result of the stay vane region, and only one point is different, namely, the STFT result (figure 12) can determine that the time for entering or being positioned at the continuous high-frequency component d with the higher amplitude of about 100-250 Hz of the measuring point sv1 of the stay vane region is earlier than or longer than the time period of 8.5-10 seconds, which indicates that the vibration source of the continuous high-frequency component with the higher amplitude of about 100-250 Hz is also in a region closer to the stay vane. The continuous low frequencies a and b with higher amplitude of about 5-50 Hz existing between 4-7 seconds almost disappear, and the following analysis can be determined: the excitation source of these low frequency components is located near the bladeless region.

And (3) pressure signal time-frequency analysis of monitoring points at the inlet of the runner: the time domain graph of the pressure signal at the wheel inlet monitoring point rn1 thus obtained is calculated, and FFT is performed on the time domain graph to analyze the frequency components of the wheel inlet pressure pulsation, and the result is shown in fig. 13. The same as the bladeless area, three high-frequency bands with high amplitudes of 56-73 Hz, 125-150 Hz and 200-210 Hz and a continuous low-frequency band with extremely high pulsation amplitude of 0-50 Hz which are caused by dynamic and static interference of the rotating wheel exist in the pressure signal at the inlet of the rotating wheel, but the pulsation amplitudes of the high-frequency bands are higher than that of a measuring point in the bladeless area. Furthermore, as shown in fig. 12, any frequency value in the range of 0 to 250Hz, which can be determined, existing at the measuring point of the runner inlet is generally much higher than the pulsation amplitude of the corresponding frequency at other places in the turbine, which is related to the special complex flow at the runner inlet during the load shedding process. The STFT analysis result (FIG. 14) is substantially the same as the time-frequency characteristic distribution of the pressure of the measuring point in the bladeless region, and the frequency value of any frequency in the range of 0-250 Hz which can be distinguished at the measuring point at the inlet of the runner is generally much higher than the pulsation amplitude of corresponding frequency at other positions in the water turbine as the FFT analysis result.

And (3) pressure signal time-frequency analysis of the inlet of the draft tube: the time domain diagram of the pressure signal of the draft tube inlet measuring point dt1 is calculated, and in order to analyze the time-frequency characteristic of the draft tube inlet pressure pulsation, the pressure signal of the measuring point dt1 is subjected to FFT and STFT, and the result is shown in FIG. 15 and FIG. 16. Compared with a bladeless area, the pressure signal at the position of a measuring point dt1 of the draft tube inlet comprises frequency bands with slightly higher pulse amplitudes of about 56-73 Hz, 113-146 Hz and 168-212 Hz caused by dynamic and static interference of the rotating wheel, but the pulse amplitude of the pressure band is lower than that of the bladeless area, which shows that the flow at the outlet of the rotating wheel is better than that at the inlet of the rotating wheel. The vibration source also comprises a group of continuous low frequencies a and b with higher amplitude of 0-50 Hz, the pulse amplitude of the continuous low frequencies a and b is lower than that of a bladeless area, and the vibration source of the continuous low frequencies of 0-50 Hz is slightly far away from the inlet of the draft tube from the bladeless area. And different from the bladeless area, continuous high frequencies c and d with higher pulse amplitude of 50-250 Hz similar to those of the volute area, the fixed guide vane area and the movable guide vane area exist at the inlet of the draft tube within the time period of 6.5-8.5 seconds, and only slightly deviate in the existing time and frequency range, which shows that except the bladeless area, vibration sources causing the continuous high frequencies with higher pulse amplitude of 50-250 Hz exist at the two sides of the upper and lower streams of the draft tube. And the pulsation amplitude of the frequency bands of 9 times, 18 times and 27 times of the frequency conversion within the time period of 6.5-8.5 seconds is obviously higher than the condition of the same time period of the bladeless area, which is the result of dynamic and static interference at the inlet of the draft tube and continuous high-frequency coupling action of 50-250 Hz. Meanwhile, the inlet of the draft tube is slightly provided with two frequency bands similar to the bladeless area and opposite to the 18-time and 27-time rotating frequency waveforms, the pulse intensity of the two frequency bands is weaker than that of the bladeless area, and therefore the following can be determined: these frequency bands opposite to the blade frequency and its harmonic waveform are definitely related to the dynamic and static interference of the rotating wheel.

Pressure signal time-frequency analysis of the draft tube straight cone section monitoring points: and calculating a time domain diagram of the obtained pressure signal of the straight-tapered section measuring point dt5 of the draft tube, and performing FFT and STFT on the pressure signal of the measuring point dt5 by comparing with the inlet measuring point of the draft tube, wherein the obtained FFT and STFT results are shown in fig. 17 and fig. 18. It can be seen from the figure that since dt5 is farther away from the wheel than dt1, the influence of the dynamic and static interference is small, and the frequency band with weaker pulse intensity is represented by 18 times and 9 times of frequency. And the pulse intensity of the low-frequency components a and b existing in 4-7 seconds and having the frequency of 0-50 Hz is continuously weakened along with the increase of the distance from the bladeless area. However, the pulse amplitude intensity of the c frequency component existing in the vicinity of 7.5 seconds and the d frequency component existing in 8.5-10 seconds is enhanced compared with that of the dt1 of the draft tube inlet, and the existence time is later than dt 1. Furthermore, similar to volute region sp1, stay vane region sv1, moving vane region gv1, vaneless region vl1 and draft tube inlet dt 1.

Pressure signal time-frequency analysis of the elbow section monitoring point of the draft tube: and calculating a time domain graph of the obtained pressure signal of the measuring point dt9 of the elbow section of the draft tube, and performing FFT and STFT on the pressure signal of the measuring point dt5, wherein the results are shown in FIGS. 19 and 20. According to the result, it can be determined that most of analysis arguments of the elbow section measuring point dt9 of the draft tube and the straight taper section measuring point dt5 of the draft tube are the same, only a high frequency band within a range of 50-250 Hz has a large difference, continuous high frequency components c and d of the elbow section measuring point dt9 close to the downstream of the draft tube represent two more obvious (compared with other 7 measuring points) high-amplitude pulse frequency bands within a time period of 7-8 seconds and 8-10 seconds, and the forms of the continuous high frequency band within the time period of 8-10 seconds and the continuous frequency band within the time period of 50-250 Hz within the time period of 7-8 seconds are similar.

Fifthly, the dynamic characteristic parameters comprise external characteristics of flow and runner torque and comprehensive external characteristics;

for flow and rotor torque out characteristics: and analyzing the time-varying process of the flow (figure 21) of the pump turbine and the moment (figure 22) of the runner in the load shedding process according to the characteristic distribution of the pressure pulsation time-frequency characteristics. As can be seen from the figure, within 4-6 seconds of the continuous low-frequency band a with higher pulsation amplitude of pressure pulsation, obvious pulsation occurs in the flow rate of the pump turbine and the external characteristics of the rotating wheel moment. Analyzing the flow and the wheel torque from the external characteristics to generate the pulsation characteristics is that the wheel torque just passes near a zero point in the time period (4-6 seconds), and the rotating speed of the wheel in the time period is generally higher. In the process that the moment of the rotating wheel changes from a positive value to a negative value, the water pump turbine is switched from the acceleration process of the water turbine to the braking process, the retardation effect of the rotating wheel on the water flow in the rotating wheel is suddenly changed into a pushing effect, and in the process of switching the working condition of the discontinuity, the water flow is considered to have certain inertia, so that the water flow in a flow channel of the water turbine can generate a strong water hammer impact oscillation effect inevitably, and meanwhile, the water head of the water turbine can reach the peak value of the load shedding process. And corresponding to a continuous frequency band b with a higher pulsation amplitude in about 6.5-7 seconds in pressure pulsation time-frequency analysis, the flow characteristic shows that the flow of the pump turbine is changed from a positive value to a negative value in about 6.5-7 seconds, the pump turbine enters a reverse pump working condition from a turbine braking working condition, and the pump turbine undergoes a second abrupt working condition conversion process in a load shedding process. At the moment, the water flow at the downstream of the draft tube flows back in a sudden reverse direction to form a large reverse impact effect on the pump turbine, and simultaneously, the water head of the pump turbine is rapidly reduced to the lowest value. In the first 0.8 seconds of the load shedding process, the guide vanes are quickly closed, so that the pushing torque of the flow and the water flow to the rotating wheel is quickly reduced, the rising of the rotating speed of the unit is effectively controlled, and the pressure pulsation is small in the first few seconds of the load shedding process.

For the comprehensive external properties: unit flow rate Q₁₁Unit speed n₁₁And unit moment T₁₁The equal unit parameters can be comprehensive characteristic indexes which can comprehensively reflect the running conditions of the pump turbine, and Q of numerical value prediction is realized for comprehensively evaluating the running conditions of the pump turbine in the load shedding process at multiple angles₁₁-n₁₁And T₁₁-n₁₁The overall characteristic curves are shown in FIGS. 23 and 24, which are shown at T₁₁Equal to zero and Q₁₁The Q of the pump turbine is equal to zero and is near the critical point of the conversion from the working condition of the water turbine to the braking working condition and then to the working condition of the reverse pump₁₁、n₁₁And T₁₁The general trend of the running track of the comprehensive characteristic parameters repeatedly fluctuates, and the severe low-amplitude high-frequency pulsation is accompanied, so that the running condition of the water pump water turbine is extremely unstable when the water pump water turbine is subjected to the critical point of the two working condition conversion in the load shedding process. The pressure pulsation time-frequency distribution characteristics of continuous frequency bands with higher pulsation amplitudes of a (about 4.5-6 seconds) and b (about 6.5-7.5 seconds) of the pressure pulsation signals of the water turbine of the water pump in the load shedding process are further explained more comprehensively from the perspective of comprehensive external characteristics.

Step six, analyzing the vorticity evolution process as follows: from the vorticity distribution of the flow field in the pump turbine during load shedding (fig. 25-30) it can be determined: in the first 3.8 seconds, the flow state of the internal flow field of the water pump turbine is smooth, only a small amount of vortexes with weak front degrees are arranged in the inlet area of the fixed guide vane and the rotating wheel, and the pressure and the external characteristics of the water pump turbine in the time period do not fluctuate too much; the load shedding process is carried out for 3.82 seconds, and two vortexes tangential along the circumference are formed near the bladeless area at the inlet sides of the two blades of the rotating wheel due to the high-speed shearing of the rotating wheel area; then the flow velocity of the bladeless area is continuously increased, so that the velocity gradient between the bladeless area and the rotating wheel is continuously increased, a plurality of tangentially distributed vortexes are also formed near the inlets of other blades under the high-speed shearing action of the inlet of the rotating wheel, the strength of the tangentially distributed vortexes is continuously increased along with the time advance until the load process is about 5.1 seconds, the water moment of the rotating wheel is zero at the moment, the rotating speed of the rotating wheel reaches the maximum runaway rotating speed, the tangentially distributed vortexes are developed to almost fill the whole bladeless area, and then the strength of the vortexes is also continuously increased along with the continuous increase of the flow velocity of the bladeless area; when the load shedding process is carried out for about 6.44 seconds, the water pump turbine is positioned near a zero flow working condition, and the water pump turbine gradually enters a reverse water pump working condition from the working condition of the water turbine, so that the vortex of the bladeless area gradually enters the double-row blade grid along with reverse incoming flow under the impact action of reverse incoming flow of a downstream draft tube, the vortex quantity distribution of the bladeless area is gradually reduced until about 7 seconds, the reverse flow of the water pump turbine reaches the maximum value, and the distributed vortex of the bladeless area is thoroughly flushed out of the double-row blade grid; then, as the reverse flow of the pump turbine is gradually reduced, the flow velocity of the bladeless area is also gradually reduced, but under the shearing action of the rotating wheel area, 9 tangential distributed vortexes are formed at the interface of the rotating wheel and the bladeless area again, the distributed vortexes continuously develop along with the advancing time and reach about 9.4 seconds, at the moment, the 9 tangential distributed vortexes near the reverse zero flow working condition are gradually connected into an annular whole, and the upstream and the downstream of the turbine are completely separated at the interface of the rotating wheel and the bladeless area. Just in the process from the working condition of the water turbine to the working condition of the brake and then to the working condition of the water pump after load shedding, the vortex quantity change process exists in the internal flow field of the water pump turbine, so that high-amplitude continuous low-frequency pulsation represented by a frequency component appears in a pressure pulsation frequency spectrogram of the water pump turbine near a zero moment working condition point, the repeated fluctuation of the overall trend is shown on the comprehensive external characteristic, and the dynamic unstable characteristic of the severe low-amplitude high-frequency pulsation is accompanied; a high-frequency high-amplitude component represented by a c frequency component appearing in a pressure pulsation frequency spectrum diagram of the pump turbine near a maximum reverse flow operating point; the continuous frequency band represented by the frequency component d appears in the pressure pulsation frequency spectrum diagram of the pump turbine near the working point of the reverse flow near the zero flow.

Sixthly, the analysis process of the flow state in the tail water pipe is as follows: so far it can be determined from the previous analysis: in the load shedding process, a frequency component in a 4.5-6 second time period of a pressure signal spectrogram of the water pump water turbine is caused by a tangential distribution vortex which is formed near a bladeless area by high-speed shearing between a runner and the bladeless area and moves along the circumferential direction; the frequency b component band of about 6.5-7.5 seconds is caused by the instantaneous impact action of reverse counter flow on a water pump turbine after the water pump enters a working condition; and c frequency components existing after about 7.5 seconds exist at other measuring points upstream and downstream of the bladeless area, but do not exist in the bladeless area, and the vorticity distribution and streamline distribution of the bladeless area after 7.5 seconds can be determined, wherein the upstream and downstream of the bladeless area are almost blocked by high-speed circulation ("water blocking ring") in the bladeless area, so that the c frequency components upstream and downstream of the bladeless area are determined to be caused by different vibration sources. From the foregoing vorticity analysis, it is found that the c frequency component at each measurement point upstream of the bladeless region is caused by the vorticity of the flow vortex and the span vortex in the double row cascade. And the excitation source of the c frequency component of the pressure signal at each point dt1, dt5 and dt9 at the downstream of the bladeless region needs to be determined by analyzing the flow state of the draft tube. According to the streamline distribution of the draft tube at each moment in the load shedding transition process (fig. 31-34), it can be determined that: the vortex motion in the draft tube in the first 6 seconds is mainly concentrated in the central area of the draft tube, and the wall surface of the draft tube is separated by a layer of pre-vortex flow flowing out from the runner, so that the vortex motion in the central area of the draft tube has small influence on the points dt 1-dt 9 close to the wall surface. And after 6 seconds, the pump turbine enters a working condition of a reverse pump, a series of vortexes are formed near the wall surface of the straight conical section due to high-speed shearing action between reverse backflow of the central area of the straight conical section of the tail water pipe and downward pre-rotational flow of the area near the wall surface of the tail water pipe flowing out from the rotating wheel, and periodic action is performed on the wall surface (as can be seen from the figure, the vortexes are obviously far away from dt1 and are transmitted to dt 5). In the elbow section, due to the action of centrifugal force, reverse backflow flows along the outer side of the elbow section and interacts with forward flow measured in the elbow section, so that a large vortex is formed near the elbow section, and the vortex flowing down from the straight cone section jointly form periodic action on the wall surface of the elbow section. Just because of the special flow in the draft tube in the load shedding process, a high-frequency band represented by a c frequency component appears in a pressure signal frequency spectrum diagram of each point dt 1-dt 9 in the draft tube after the load shedding of the pump turbine is about 7.5 seconds.

The analysis method firstly carries out the boundary condition setting and the calculation setting of CFD calculation in the load shedding transition process, then carries out detailed time-frequency characteristic analysis on the pressure pulsation of different component measuring points in the whole flow domain, obtains four special frequency components a, b, c and d, and preliminarily analyzes the positions of excitation source of the four special frequency components. Then, the incentive of the special frequency component is given from the perspective of the external characteristic. And finally, an internal flow field in the load shedding transition process is analyzed, and specific reasons for forming special frequency components are given from two aspects of vorticity and flow state of a draft tube.

Although the present invention has been described with reference to the preferred embodiments, it should be understood that various changes and modifications can be made therein by those skilled in the art without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A dynamic characteristic analysis method for a pump turbine in a load shedding process is characterized by comprising the following steps:

2. The dynamic behavior analysis method of claim 1, wherein the boundary condition, calculation setting and turbulence model setting process of step two comprises:

3. The dynamic characteristics analysis method of claim 1, wherein the calculation process of the rotating speed of the rotating wheel in the third step comprises:

wherein, M is the resultant moment (N.m) borne by the rotor;

j-rotor moment of inertia (kg. m)²)；

ω -rotor angular velocity (rad/s);

t-time(s);

4. The dynamic characteristics analysis method according to claim 1, wherein the arrangement positions of the pressure measurement points in step four include: in a plane passing through the middle height of the double-row cascade, volute measuring points SC-MP-1-4, fixed guide vane inlet measuring points sv-1-6, movable guide vane inlet measuring points gv-1-20, bladeless area measuring points vl-1-20 and rotating wheel inlet measuring points rn-1-9 are respectively arranged along the clockwise direction: in a plane passing through the movable guide vane and close to the height of the bottom ring, fixed guide vane inlet measuring points sv-dn-1-6, movable guide vane inlet measuring points gv-dn-1-20, bladeless area measuring points vl-dn-1-20 and runner inlet measuring points rn-dn-1-9 are respectively arranged along the clockwise direction: the height of the transitional guide vane close to the top cover is respectively provided with fixed guide vane inlet measuring points sv-up-1-6, movable guide vane inlet measuring points gv-up-1-20, bladeless area measuring points vl-up-1-20 and rotating wheel inlet measuring points rn-up-1-9 in the clockwise direction: monitoring points DT 1-12 are respectively arranged on three different sections of the draft tube.

5. The dynamic characteristic analysis method according to claim 3, wherein the concrete step of obtaining the pressure pulsation characteristic of the pump turbine in the load shedding process in the fourth step is: