RU2229708C2 - Process detecting flaws, predominantly corrosion defects, in pipe-lines - Google Patents

Abstract

FIELD: nondestructive inspection of pipe-lines, instrumentation. SUBSTANCE: process detecting flaws in pipe-lines consists in placement of pair of acoustic transducers on ends of tested section of pipe-line. One transducer in each pair records longitudinal oscillations and another transducer registers lateral vibrations. Oscillations generated by flaw are recorded by all transducers simultaneously with their subsequent filtration to isolate signals with frequency of 1000-2500 Hz; signals of acoustic transducers in each pair are compared with one another to extract signals from flaws. Signals from transducers of longitudinal oscillations generated by flaw are converted to digital codes, high-frequency phase oscillations are modulated by digital code from one transducer, phase-shifted signal is power amplified, emitted into the air, detected at another end of tested section of pipe-line, multiplied and divided by phase in two, harmonic voltage is extracted, its initial phase is stabilized, multiplied with phase-shifted signal, modulating digital code is derived, delayed in time, multiplied with another digital code, cross correlation function of digital codes is isolated, delay time is varied till maximal value of cross correlation function is achieved, maximal value of cross correlation function is maintained by automatic change of delay time, delay time is recorded, position of flaw in tested section of pipe-line is established and recorded. EFFECT: enhanced productivity of process, increased length of sections of pipe-line diagnosed reliably. 2 dwg

Description

The proposed method relates to a control and measuring technique and can be used to detect defects in pipelines, mainly corrosion defects in central heating pipelines, hot and cold water supply of municipal utilities in cities.

Corrosion defects include:

- local thinning of the pipe wall to a thickness of less than 50% of the nominal;

- a series of ulcers covering an area with an axial strike of more than 100 mm;

- defects in welds;

- opening of cracks through which there is a flow of water of various, including low, intensity and others.

Known methods for detecting defects in pipelines (USSR author's certificate No. 380909, 411268, 642575, 934269, 1216551, 1283566, 1610347, 1657988, 1672105, 1679232, 1705709, 1733837, 1777014, 1778597, 1812386; Russian patents No. 21353787, 213; No. 4289019, 4570477, 5038614; UK patent No. 1349129; French patent No. 2498325; Japanese patents No. 59-38537, 60-24900, 63-22531; Pipeline transport of liquid and gas. M., 1993 and others).

Of the known methods, the closest to the proposed one is the “Method for the detection of corrosion defects in water supply pipelines” (RF patent No. 2138037, G 01 N 29/04, 1998), which is selected as a prototype.

This method can be used to monitor the status of central heating pipelines, hot and cold water supply of municipal utilities in cities. The simplification of the process of determining the location of a defect and its nature, as well as the independence of the determination process from the material properties of the controlled medium, is achieved by the fact that pairs of acoustic sensors are installed at the ends of the studied section of the pipeline, one of which records longitudinal vibrations in each pair and the other transverse vibrations. After processing the signals from the sensors, the location and nature of the detected defect is determined.

The disadvantages of this method are the low productivity due to the need to lay the cable between the sensors installed at the ends of the monitored section of the pipeline, and the impossibility of its use in a number of situations, for example, cable crossings are not used at the crossings through transport routes.

An object of the invention is to increase productivity and increase the lengths of reliably diagnosed sections of pipelines of central heating, hot and cold water supply of municipal utilities in cities.

The problem is solved in that according to the method for detecting defects in pipelines, according to which pairs of acoustic sensors are installed at the ends of the monitored section of the pipeline, longitudinal vibrations are detected with one sensor and transverse vibrations with the other, while all the vibrations generated by the defect are recorded by all sensors simultaneously with their subsequent filtering to extract signals with a frequency of 1000-2500 Hz and compare the signals of acoustic sensors in each pair with each other to highlight I signals from the defect, the signals from the longitudinal oscillation sensors generated by the defect are converted into digital codes, the digital code of one of the sensors manipulates the high-frequency oscillation in phase, amplifies the received phase-manipulated signal by power, emits it into the air, receives the phase-manipulated signal at the other end of the controlled area pipelines, multiply and divide its phase into two, isolate the harmonic voltage, stabilize its initial phase, multiply with the received phase-manipulated signal, isolate they modulate the digital code, delay it in time, multiply it with another digital code, select the cross-correlation function of the digital codes, change the delay time until the maximum value of the cross-correlation function is obtained, maintain the maximum value of the cross-correlation function by automatically changing the delay time, fix the delay time, determine the location of the defect in the controlled section of the pipeline and register it.

The structural diagram of a device that implements the proposed method is presented in figure 1.

Timing diagrams explaining the principle of operation of the device shown in figure 2.

The device contains a pipeline 1, defect 2 and pairs of acoustic sensors 3.1 and 4.1, 3.2 and 4.2. In this case, longitudinal vibrations are detected by sensors 3.1 and 3.2, and transverse vibrations generated by defect 2 by sensors 4.1 and 4.2. Corrosion defects predominate as defects, which include:

- local thinning of the pipe wall to a thickness of less than 50% of the nominal;

- a series of ulcers covering an area with an axial strike of more than 100 mm;

- defects in welds;

- opening of cracks through which there is a flow of water of various, including low, intensity and others.

The water flow (pressure wave) affects the place of thinning of the pipe wall, which causes the forced vibrations of the pipe (membrane effect) and the generation of the corresponding longitudinal wave, which propagates through the water with the corresponding velocity V = 1200 m / s. Fluctuations from thinning of the pipe wall, as a rule, are pulsed in nature with a frequency in the range of 1000-2500 Hz. The interaction of the longitudinal pressure wave from defect 2 with the pipe wall at a specific point is equivalent to the well-known solution to the problem of the action of a concentrated dynamic force on the shell and, in accordance with the solution, the longitudinal movements of the pipe wall over the transverse ones predominate.

In addition, the pipe is under the influence of various kinds of vibrations associated with the flow of water and its structures. These oscillations mainly cause the appearance of standing waves and Rayleigh waves (waves of transverse displacements).

Thus, at the point of sensor installation, the longitudinal and transverse displacements will be due to the superposition of the two main waves, and the longitudinal displacements will have in amplitude more components associated with the radiation of the wave from the defect than the transverse ones. Sensors 3.1, 3.2, 4.1, 4.2 convert acoustic vibrations into electrical signals. Amplifiers of electrical signals 5.1, 5.2, 6.1, 6.2 and filters 7.1, 7.2, 8.1, 8.2 are connected to the sensors in series, which provide the selection of signals in the frequency range of 1000-2500 Hz. The comparison unit 9.1 (9.2) and the switch 10.1 (10.2) and the analog-to-digital converter 11.1 (11.2) are connected in series to the filters. To the output of the analog-to-digital converter 11.2, a phase manipulator 13 is connected in series, the second input of which is connected to the output of the high-frequency generator 12, a power amplifier 14 and a transmitting antenna 15. A high-frequency amplifier 17, a phase doubler 18, and a divider 19 are connected in series to the output of the receiving antenna 16 two phases, a narrow-band filter 20, a frequency detector 21, a trigger 22, a balance switch 23, the second input of which is connected to the output of the narrow-band filter 20, a phase detector 24, the second input of which is connected to the second the output of the high-frequency amplifier 17, an adjustable delay unit 26, a multiplier 27, the second input of which is connected to the output of the analog-to-digital converter 11.1, a low-pass filter 28, a correlation function level indicator 29, an extreme regulator 30, an adjustable delay unit 26, a time delay indicator 31 computing unit 32 and registration unit 33. The adjustable delay unit 26, the multiplier 27, the low-pass filter 28, the correlation function level indicator 29 and the extreme controller 30 form a correlator 25.

A method for detecting defects is as follows.

The method is intended to detect and determine the location of corrosion defects in pipelines with a diameter of more than 80 mm, having a wall thickness of 5-15 mm, through which the product or coolant is pumped, in particular water.

The method is based on the phenomenon of emission of acoustic frequency vibrations (emissions) when a water current (dynamic pressure) acts on these defects.

Vibroacoustic signals excited by defect 2 propagate along pipeline 1 in both directions and are sensed by sensors 3.1, 4.1 and 3.2, 4.2, from the outputs of which electric signals are fed to amplifiers 5.1, 6.1 and 5.2, 6.2, respectively. With the help of filters 7.1, 8.1 and 7.2, 8.2, the working frequency band (1000-2500 Hz) is allocated.

In order to isolate the signals from defect 2 against the background of pipe vibrations as a whole, autocorrelation is carried out according to the records of longitudinal and transverse vibrations, and then the results of autocorrelation of these records are compared in comparison blocks 9.1 and 9.2.

If there is a defect in the controlled section of the pipeline, the comparison blocks 9.1 and 9.2 are activated and form constant voltages, which are supplied to the control inputs of the keys 10.1 and 10.2, opening them. In the initial state, keys 10.1 and 10.2 are always closed. In this case, electrical signals corresponding to longitudinal acoustic vibrations generated by defect 2, through the public keys 10.1 and 10.2, are fed to analog-to-digital converters 11.1 and 11.2 and converted into digital codes. From the output of the first analog-to-digital converter 11.1, the digital code is fed directly to the first input of the correlator 25. Due to the large spatial distance of the second analog-to-digital converter 11.2 from the correlator 25, a radio channel is used to transmit a digital code from the output of the analog-to-digital converter 11.2 to the second input of the correlator 25.

For this purpose, a high-frequency oscillation is generated by the generator 12 (Fig. 2, a)

U _c (t) = U _c cos (ω _c t + φ _c ),

0≤t≤T _c ,

where U _c , ω _c , φ _c , T _c is the amplitude, carrier frequency, initial phase, and duration of the high-frequency oscillation, which arrives at the first input of the phase manipulator 13, the second input of which is supplied with the modulating code M (t) (Fig. 2, b) from the output of the second analog-to-digital converter 11.2. At the output of the phase manipulator 13, a phase-shift (PSK) signal is generated (Fig. 2, c).

U ₁ (t) = U _c · cos [ω _c t + φ _k (t) + φ _s ],

0≤t≤T _c ,

where φ _k (t) = {0, π} is the manipulated phase component that displays the phase manipulation law in accordance with the modulating code M (t), and φ _k (t) = const for Kτ _e <t <(K + 1) · Τ _e and can change abruptly at t = Kτ _e , i.e. at the boundaries between elementary premises (K = 0, 1, 2, ... N-1);

τ _e , N - the duration and number of chips that make up a signal of duration T _s (T _s = N · τ _e ).

This signal, after amplification in the power amplifier 14, is transmitted by the transmitting antenna 15 to the ether, received by the receiving antenna 16, and after amplification in the high-frequency amplifier 17, is fed to the first input of the phase detector 24, which is a series multiplier and a low-pass filter. At the same time, the PSK signal is fed to the input of the phase doubler 18, which can be made in the form of a multiplier, the same PSK signal is fed to two inputs of which. At the input of the phase doubler 18, a harmonic voltage is generated (Fig. 2, g)

U ₂ (t) = U ₂ cos (2ω _c t + 2φ _s ),

0≤t≤T _c ,

where U ₂ = 0.5K ₁ · U $\genfrac{}{}{0ex}{}{\text{2}}{\text{c}}$ ;

K ₁ - transmission coefficient of the multiplier, in which phase manipulation is already absent. This voltage is fed to the input of the phase divider 19 into two, the output of which is formed by the following harmonic voltage (figure 2, d)

U ₃ (t) = U ₂ cos (ω _c t + φ _c ),

0≤t≤T _c .

The initial phase of the obtained harmonic voltage can have two stable values: φ _c and φ _c + π. This is easy to show analytically. If you make a division similar to the previous one, but previously adding an angle of 2π to the argument, which does not change the initial voltage, then after dividing the phase by two, you get the voltage shifted in phase by π

U ₃ '(t) = U ₂ · cos [(2ω _c t + 2φ _c + 2π) / 2] = U ₂ · cos (ω _c t + φ _c + π).

Therefore, the ambiguity of the initial phase of the obtained harmonic voltage follows from the fission process itself. The physically indicated ambiguity of the initial phase is explained by the unstable operation of the phase divider 19 into two. This phenomenon is called “reverse operation” and is inherent in all known devices (Pistolkoria A.A., Siforova V.I., Kostala D.F., Travina G.A.), which emit the reference voltage necessary for the synchronous detection of FMN signals directly from the received QPSK signal.

The phenomenon of “reverse work” is due to spasmodic transitions of the initial phase of the reference voltage from one state φ _s to another φ _s + π under the influence of noise, short termination of reception, and other factors. These transitions during the reception of the QPSK signal occur at random times (for example, t ₁ , t ₂ ) (figure 2, e). At the same time, at the output of the phase detector 24, a distorted analog of the modulating code M ₁ (t) is extracted (Fig. 2, e), which reduces the reliability of determining the location of the defect 2.

To stabilize the phase of the reference voltage and eliminate the phenomenon of “reverse operation”, a desktop detector 21, a trigger 22, and a balance switch 23 are used.

When the initial phase of the reference voltage jumps by 180 ° at time t ₁ (Fig. 2, e), a positive short pulse is generated at the output of frequency detector 21, and when the initial phase jumps by -180 ° at time t ₂ (return of the initial phase to the initial state) is a negative impulse (figure 2, g). Alternating short pulses from the output of the frequency detector 21 (FIG. 2, g) control the operation of the trigger 22, the output voltage of which (FIG. 2, h), in turn, controls the operation of the balance switch 23.

In a stable state, when the initial phase of the reference voltage coincides, for example, with the zero phase of the received PSK signal, a negative voltage is generated at the output of the trigger 22 and the balance switch 23 is in its initial position, at which the reference voltage is supplied from the output of the narrow-band filter 20 to the reference the input of the phase detector 24 without change.

When an abrupt change in the initial phase of the reference voltage by + 180 °, caused, for example, by the unstable operation of the phase divider 19 into two under the influence of interference, the trigger 22 positive pulse (figure 2, h) from the output of the frequency detector 21 is transferred to another stable state. In this case, the output voltage of the trigger 22 at time t ₁ becomes and remains positive until the next phase jump at time t ₂ , which returns the initial phase of the reference voltage to its original state. The positive output voltage of the trigger 22 transfers the balance switch 23 to another stable state, in which the reference voltage from the output of the narrow-band filter 20 is supplied to the reference output of the phase detector 24 with a change in the initial phase by -180 °. This eliminates the instability of the initial phase of the reference voltage and the associated “reverse work”.

Therefore, the frequency detector 21 provides detection of the moment of occurrence of “reverse operation 2, and the trigger 22 and the balance switch 23 eliminate it.

In this case, the reference voltage with a stable initial phase (Fig. 2, and) is supplied to the reference input of the phase detector 24.

U ₄ (t) = U ₂ · cos (ω _c t + φ _s ),

0≤t≤T _c .

At the output of the phase detector 24, a low-frequency voltage is generated (figure 2, k)

U _n (t) = U _n cosφ _k (t),

0≤t≤T _c ,

where U _n = 0,5K ₂ · U _c · U ₂ ;

To ₂ - the gain of the phase detector, proportional to the modulating code M (t).

This voltage is supplied to the second input of the correlator 25, consisting of an adjustable delay unit 26, a multiplier 27, a low-pass filter 28, a correlation function level indicator 29 and an extremal regulator 30. The correlation function R (τ) obtained at the output of the correlator 25, measured by the indicator 29 , has a maximum value when the introduced adjustable delay τ is equal to the transport delay of the signal of one sensor with respect to the signal of another sensor. The extreme controller 30 maintains the value of R (τ) at the maximum level by acting on the adjustable delay unit 26. The scale 31 of the adjustable delay unit 26 is graded directly in the values of the difference in the time of arrival of longitudinal waves at the ends of the monitored section of the pipeline 1.

Having determined the time difference of arrival of longitudinal waves (t ₁ , t ₂ ) generated by defect 2 at the ends of the monitored section of length R (Fig. 1), the location of defect 2 is determined in computing unit 32

L = 0.5 [R (1-V _sr / s) - (s ² -V $\genfrac{}{}{0ex}{}{\text{2}}{\text{wed}}$ ) / s · (t ₂ -t ₁ )],

where c is the speed of acoustic waves;

V _cp is the average speed of the transported product (water, oil, gas, etc.), which is recorded by the registration unit 33.

Thus, the proposed method in comparison with the prototype and other technical solutions for a similar purpose provides increased productivity and increased lengths of reliably diagnosed sections of central heating pipelines, hot and cold water supply of municipal utilities. This is achieved by using the radio channel and the correlation processing of the used PSK signals. These signals allow the use of a new type of selection - structural selection. This means that there is a new possibility of detection and separation. Signals operating in the same frequency band and at the same time intervals.

Complex QPSK signals also have high energy and structural secrecy.

To isolate the modulating code M (t) from the received QPSK signal, its synchronous detection is used. Moreover, the reference voltage required for synchronous detection of the received PSK signal is extracted directly from the received PSK signal. And the phenomenon of “reverse operation” arising from this is eliminated by stabilizing the initial phase of the reference voltage, which is carried out using a frequency detector, trigger, and balance switch.

Claims (1)

A method for detecting defects in pipelines, mainly corrosion defects in water supply pipelines, by which pairs of acoustic sensors are installed at the ends of the pipeline section to be monitored, longitudinal vibrations are recorded with one sensor and transverse vibrations with the other, and all oscillations generated by the defect are recorded by all sensors simultaneously with their subsequent filtering to extract signals with a frequency of 1000-2500 Hz and compare the signals of acoustic sensors in each pair m I am waiting for myself to isolate signals from defects, characterized in that the signals from the longitudinal vibration sensors generated by the defect are converted into digital codes, the high-frequency phase oscillation is manipulated with the digital code of one of the sensors, the received phase-manipulated signal is amplified by power, emitted, transmitted phase-shifted signal at the other end of the monitored section of the pipeline, multiply and divide its phase into two, isolate the harmonic voltage, stabilize its initial phase, multiply with with an accepted phase-manipulated signal, a modulating digital code is extracted, it is delayed by time, it is multiplied with another digital code, the mutual correlation function of the digital codes is extracted, the delay time is changed until the maximum value of the mutual correlation function is obtained, the maximum value of the mutual correlation function is automatically changed by the delay time, fixed delay time, determine the location of the defect in a controlled section of the pipeline and register it.

Images