GB2191860A

GB2191860A - Method of detecting an obstruction or a discontinuity in a tube

Info

Publication number: GB2191860A
Application number: GB08614957A
Authority: GB
Inventors: Eric Sidney Morgan
Original assignee: Central Electricity Generating Board
Current assignee: Central Electricity Generating Board
Priority date: 1986-06-19
Filing date: 1986-06-19
Publication date: 1987-12-23
Anticipated expiration: 2006-06-19
Also published as: GB2191860B; GB8614957D0

Abstract

The method comprises applying an acoustic pulse signal including a predetermined wide range of frequencies to the tube 10 and analysing the frequency spectrum of the signal reflected from any obstruction 11 or discontinuity 12 in the tube thereby to characterise that obstruction or discontinuity. The method can be carried out automatically and can distinguish between holes and simple increases in diameter in the tube and can provide an indication of the shape of any obstruction discovered. The apparatus required includes a signal generator 16, a loudspeaker 14, microphone 15, a filter 18 and a computer 17 for analysing the spectrum of the reflections. <IMAGE>

Description

SPECIFICATION Method of detecting an obstruction or a discontinuity in a tube This invention relates to a method of detecting an obstruction or a discontinuity in a tube.

In a GB-A-1482592 there is described such a method for use, for example, with tubes in steam raising boilers as used in electricity power generating stations, in which an acoustic pulse signal is applied to one end of a tube for propagation along the tube and reflection from any obstruction or discontinuity in the tube, any reflected signal being detected at the one end of the tube and displayed on a cathode ray tube.

The cathode ray tube has a time scan synchronised with the applied signal, and thus the display will show the time between application of a pulse and detection of its reflection, this giving a measure of the distance to the detected obstruction on discontinuity.

As described, the display is visually monitored by an operator who can distinguish between an obstruction and a discontinuity (hole) in that for an obstruction the reflected signal comprises a positive-going deflection followed by a negative-going deflection, whereas for a discontinuity or hole the reflected signal comprises firstly a negative-going deflection followed by a positive-going deflection.

While this known method is adequate for many circumstances it is wholly dependent upon the skill of the operator, and thus mistakes can be made. Further, the display gives little information about an obstruction or discontinuity (or hole) detected, giving only its position and basic nature.

In particular the method is unable to distinguish between holes and simple increases in the tube diameter, caused for example by erosion, since both give a reflected signal which is initially a negative-going deflection. Further, the method is unable to provide an indication of the shape of any obstruction discovered.

According to the invention there is provided a method of detecting an obstruction or a discontinuity in a tube, comprising applying an acoustic pulse signal including a predetermined range of frequencies to one end of a tube for propagation along the tube and reflection from any obstruction or discontinuity in the tube; detecting the reflected signal at said one end of the tube; and analysing the frequency spectrum of the reflected signal thereby to characterise the obstruction or discontinuity detected.

This invention will now be described by way of example with reference to the drawings, in which: Figure 1 is a diagrammatic representation of apparatus for use in carrying out the method of this invention; Figure 2 shows representation of frequency and time domain reflected signals obtained from holes in a tube using the method of this invention; Figure 3 shows representations of frequency and time domain reflected signals obtained from obstructions in a tube using the method of this invention; Figure 4 shows three diagrams illustrating a characterisation carried out using the method of this invention; and Figure 5 illustrates another characterisation carried out using the method of this invention.

Fig. 1 shows a tube 10 to be inspected, the tube being shown as having an obstruction 11 in the form of a bolt therein, and a hole 12 therein. Mounted at one end of the tube 10 is a transducer assembly 13 comprising a loudspeaker 14 and a microphone 15. The loudspeaker is fed with a pulse signal from a generator 16 which also supplies the signal to a computer 17.

Signals from the microphone 15 are fed by way of a filter 18 to the computer 17.

The pulses from the generator 16 each embrace a predetermined wide frequency spectrum and are of short duration. These pulses are transmitted by the loudspeaker 14 to propogate down the tube 10 to be reflected from any obstruction 11 or hole 12, the reflected pulses being received by the microphone 15 and fed to the computer 17.

Fig. 2 shows the frequency and time characteristics for pulses embracing a 43 to 1305Hz frequency spectrum and having a length 11 ms, as received by the microphone 15 after reflection from three different size holes 12 (0.5mm, imm and 3.5mm diameter) in the tube 10.

As shown the frequency spectra of the reflected pulses for the three holes are basically similar.

The amplitude of a reflected signal from a hole 12 is related to the radius a of the hole 12, the radius R of the tube 10, the wavelength A of the applied pulse and the thickness W of the tube wall:

where V/Vo is the amplitude of the reflected pulse as a fraction of the applied pulse.

This equation can be rewritten as:

For a small hole and thus a2A small compared with 4n:R2(W+ 1 .7a) then V/Vo - a2A, and thus the frequency spectra for all holes are basically similar as noted above.

For comparison purposes, the amplitudes shown in Fig. 2 for the three pulses have been shown at the same maximum value.

Considering now Fig. 3 this shows characteristics corresponding to those shown in Fig. 2, but deriving from an obstruction 11 in the tube 10, and in particular from rods 7.5, 15 and 23cm long.

A comparison between the frequency spectra of Figs. 2 and 3 shows that those characteristic of holes have a ratio of low frequency to high frequency content higher than that for obstructions, and thus this fact can be used by the computer 17 to ascertain whether a received reflected pulse derives from a hole or an obstruction.

As in the known methods disclosed in GB-A-1482592 the distance to the hole (or obstruction) along the tube 10 can be ascertained by the computer 17 from the time between application of a pulse by the loudspeaker 14 and reception of its reflection by the microphone 15.

Experimental evidence has shown that the amplitude of a reflected pulse depends upon the cross-sectional area along the length of any obstruction 11 in the tube 10 from which the reflected pulsed derives, when the wavelength of the pulse is relatively long as compared with the length of the obstruction along the tube 10. Thus, any obstruction 11 can be considered to be in the form of a rod having the same length as the obstruction 11 and having a uniform cross-sectional area equal to the average cross-sectional area of the obstruction 11. The signal received by the microphone 15 from such a rod will be a combination of reflections from the near and far ends of the rod.If a reflected pulse from such obstruction is described in the time domain as f(t), then the Fourier transform is:

If this signal is displaced by time to then the amplitude spectrum will remain the same but the phase of the components will alter. The Fourier transform of the new signal is obtained from that of the old by multiplying by exp(wtO).

Hence the Fourier transform of f(t-to) is exp(-jwt)F(w).

If two similar signals separated in time by +to and to from the centre of the obstruction 11 are considered, which two signals correspond to reflections from the near and far ends of the obstruction 11, then the transform of the combined signals f(t+t0)+f(t-t0) is: exp(jwtO)F(w) + exp(-jwt0)F(w) = 2coswt. F(w) (3) The frequency spectrum is now that of a single pulse modulated by a frequency related to the phase separation between the components of the two signals.

If equation (3) is divided by F(w) then the modulating waveform 2cost, is obtained. Maxima will occur at values given by wt,=nn or F=)1/2t,.

Hence, the frequency of separation between maxima is 1/2to.

F(w) is the spectrum from the near end of the obstruction and thus is equivalent to the spectrum of the reflected pulse from the near end of a long obstruction where there is no interference from a reflection from the far end of the obstruction.

The upper diagram in Fig. 4 shows, as an example, in solid line the spectrum for a reference long obstruction in the form of a rod 2 metres long, and in dashed line the combined spectrum obtained as described above from reflections from both ends of an obstruction in the form of a rod 23cm long. The middle diagram in Fig. 4 illustrates the ratio of the two spectra of the upper diagram, from which the modulating frequency discussed above is clearly seen. On taking a spectral analysis of the middle diagram the modulating frequency can be represented as shown in the lower diagram in Fig. 4, where the x-axis is in the terms of time, and by multiplying by a suitable factor this can be expressed in terms of length to give the length of the obstruction 11.

Calculations as set out above are all effected in the computer 17 which is thus able to indicate whether a hole or an obstruction has been detected, and if an obstruction the length thereof.

Equation (3) above shows that the spectrum of a reflected pulse from an obstruction is that of the applied pulse (reference signal) modulated by a frequency dependent upon the length of the obstruction. If the obstruction is of irregular cross-section, then there will be a plurality of modulating frequencies dependent upon the shape of the obstruction.

Thus, from the Fourier transform F(w)r of a reference signal, for example obtained as described above by recording the reflected pulse from the near end of a long rod obstruction in the tube, and the Fourier transform F(w)o of the reflected pulse from an obstruction: F(w)o/F(w)r gives a complex waveform dependent upon the shape of the obstruction.

Since the reference signal is in effect obtained from a uniform change in the cross-sectional area of the tube, this can be represented by a square wave having a long duration. Multiplying F(w)o/F(w)r by the complex spectra of such square wave and performing the inverse transform produces a reconstruction of the obstruction in the time domain, whereby an image showing the changes in cross-sectional area of the obstruction but not necessarily its true cross-section can be produced.

This procedure which is carried out in the computer 17 is illustrated by Fig. 5 which shows an obstruction, the signal obtained from the obstruction, a previously produced reference signal, and an image of the obstruction produced by the computer 17.

Claims

1. A method of detecting an obstruction or a discontinuity in a tube, comprising applying an acoustic pulse signal including a predetermined range of frequencies to one end of a tube for propagation along the tube and reflection from any obstruction or discontinuity in the tube; detecting the reflected signal at said one end of the tube; and analysing the frequency spectrum of the reflected signal thereby to characterise the obstruction or discontinuity detected.

2. A method as claimed in Claim 1, in which the ratio of the low to high frequency content in the frequency spectrum of the reflected signal is determined, a relatively high ratio being characteristic of a hole in the tube.

3. A method as claimed in Claim 1, in which the frequency spectrum of the reflected signal is compared with that of a reference signal characteristic of a hole in the tube.

4. A method as claimed in Claim 1, in which the length of an obstruction in the tube is calculated from a combined signal obtained from reflected signals from the ends of the obstruction, and a reference signal obtained from a reference obstruction in the tube.

5. A method as claimed in Claim 4, in which the calculations provide an indication of changes of cross-section area in the obstruction.

6. A method as claimed in any preceding claim, in which necessary calculations are carried out in a computer.

7. A method of detecting an obstruction or a discontinuity in a tube, substantially as hereinbefore described with reference to the drawings.