GB2138138A - Dynamic method for the detection of discontinuities in a body - Google Patents

Abstract

A method of detecting stress sensitive discontinuities in a body, comprises subjecting the body to an alternating or changing stress, directing pulses of ultrasonic signals into the body, detecting ultrasonic signals returning from the body, storing the said signals and comparing the stored signals corresponding to returning signals originating at one level of the alternating stress with those originating at another level of the alternating stress thereby to eliminate those return signals which do not vary in accordance with the applied stress. To perform the comparison, digitized ultrasonic returns maybe multiplied 9 by 1 for stresses above average, and by -1 for stresses below average, then added 11 to signals already stored. Discontinuities, not sensitive to stress will tend to produce a zero resultant in memory 12 whilst those sensitive to stress will give a non-zero resultant. <IMAGE>

Description

SPECIFICATION Dynamic method for the detection of discontinuities in a body The invention relates to ultrasonic testing, and in particularto the detection of discontinuities in a body, such as cracks and similarfaults.

It is known that the reflectivity to ultrasound of a crack in a body varies when a compressive load is applied to the body, so as to cause the crack to tend to close. The effect is believed to be due to variations in the area of contact between asperities on the sides of the crack. This phenomenon has been utilised as the basisofa method oftesting steel specimens for cracks, in which a specimen undertest is subjected to a cyclic stress, ultrasonic pulses are applied to the specimen, a particular echo pulse in each set of return pulses is selected, its amplitude is determined and displayed on a visual display unit.The repetition rate ofthe ultrasonic pulses is made to be slightly different from the cyclic rate of variation of the applied stress, so that a beat appears in the form of a cyclic variation in the ampiitude of the trace displayed on the screen of the visual display unit.

According to the present invention there is provided a method of detecting stress sensitive discontinuities in a body comprising the operations of subjecting the body to an alternating or changing stress, directing pulses of ultrasonic signals into the body, detecting ultrasonic signals returning from the bodyand producing echo signals related thereto, storing temporarilythe echo signals and comparing the stored echo signals corresponding to returning signals originating at one level of the alternating stress with those originating at another level of the alternating stress therebyto eliminatethose echo signalswhich do not vary in accordance with the applied stress.

Preferablymorethan one set of return signals is accumulated at each level ofthe alternating stress, and the two accumulated sets of return signals are compared.

Also according to the present invention there is provided apparatus for detecting stress sensitive discontinuities in a body subjected to a varying level of stress, comprising meansforproducing echo signals corresponding to pulses of ultrasound reflected from discontinuities within the body within discrete time intervals, means for producing signals relating to the varying levels of stress within the body, means for correlating the echo signals in each time interval with the level of stress in the body during that time interval, means for storing temporarily the correlated echo signals, and means for comparing stored signals derived from echo signals generated at one level of stress in the body with those generated at asecond level ofstressinthe bodytherebyto eliminate the component of the echo signals which is not sensitive to the level of stress within the body.

The invention will now be described, byway of example, with reference to the accompanying drawings, in which: Figure lisa block circuit diagram of a dynamic crack detection system embodying the invention.

Figure 2 is a block circuit diagram of a dynamic crack detection system for use when the alternating forces are varying in an unknown fashion; and Figure 3 is a block circuit diagram of another dynamic crack detection system embodying the invention.

Referring to Figure 1, a test specimen or other structure 1 has an alternating force F applied to one end 2 while the other end 3 is held in a massive mounting 4. The source of the Force F is not shown as it can be generated in any convenient manner. The force F can be applied as shown, directly between the ends 2 and 3 ofthe specimen 1, by flexing of the specimen 1, or in any other way. For example, the alternating force F may be one which occurs naturally, such as those due to wave motion if the specimen 1 forms part of an offshore structure. Atransducer 5 is arranged to inject pulses of ultrasound into the specimen 1, and to receive returning echo signals.

Both the transducer and any interfacing medium are such that they are not affected by the alternating stress in the specimen 1 so as to preventthe generation of any spurious stress-related signals. Also attached to the specimen lisa strain gauge 6, the purpose of which is to provide information aboutthe alternating stressing of the specimen 1. The returning echo signals are detected by a receiver circuit 7, digitized at descrete time intervals after each relevant transmitted pulse in an analogue to digital converter 8 and applied to a multiplier 9 in which they are multiplied by stress related signals from the strain gauge 6 which have been digitized in another analogue to digital converter 10.The outputfrom the multiplier 9 is applied to an adder 11 in which it is added to the previously stored contents of a memory 12 before being returned to the memory 12. The contents of the memory 12 are displayed on a display unit 13. Atimer 14 operates an ultrasonic pulse generator 15, controls the operation of the analogue to digital converter8, selects an appropriate address of the memory 1 2for each time interval and controls the release from the memory 12 of the previously stored digitized echo signals.

In the arrangement described, the received echo signals are amplified in the receiver 7, and digitized at discretetimeintervalsafterthe production of each relevant transmitted pulse by means of the analogue to digital converter 8. The time intervals, which are determined by the timer 14, are short enough to preserve the characteristics of the individual echo singals. The time intervals will hereinafter be referred to as time slots. The digitised signals in each time slot arethen correlated with the digitised strain gauge signals in the multiplier 9. The action of the multiplier 9 need only be simple. for example, it is adequate for echo signals to be multiplied by 1 for above-average stress and '1 for below-average stress.Multiplication by zero for near-zero stresses slightly improves the signal/noise ratio. It is important however, to ensure that the number of echo signals in any given time slot which is multiplied by 1 is equal to that which is multiplied by , so that, when the signals are added together, stress- insensitive signals will cancel out.

The treated signals in each time slot are added to all the signals which have previously occurred in the corresponding time slot, and the result is stored back in the same position in the memory 12. The memory 12thuscontains data corresponding to the algebraic total of all the echo signals which have occurred in each time slot.

If the echo signals are not affected by the stress variations in the specimen 1, then equal numbers of equal sized positive and negative signals will be added together and the contents of the memory will be zero, apartfrom any component due to random noise. Any signal which does vary with the stress applied to the specimen 1 will accumulate a number in the position appropriate to the position in the specimen 1 ofthe related stress-sensitive discontinuity, which will increase steadily with the number oftransmit/receive cycles which have occurred.As the noise component in the stored signals will accumulate only in propor tiontothesquare root of the numberoftransmit/ receive cycles, the signal/noise ratio can be increased indefinitely by increasing the length ofthe period of time overwhich observations are made.

lfthefrequency ofthe applied alternating force F is ofthe same arder of magnitude as the repetition frequency of the ultrasonic pulses, then it is desirable to synchronisethe pules of ultrasound to an even integral multiple ofthe frequency ofthe applied alternating force F.To enablethisto bedone,a synchronisation connection can be made between the strain gauge 6 and the timer 14.

Alternatively, the timer 14 can be arranged to control directlythe generation and application ofthe alternating force F, so thatthe changes in theforce F are synchronised with a sub-multiple ofthe ultrasonic pulse repetition frequency.

Ifthe alternating force Fvaries unpredictably, then synchronisation between the pulses of ultrasound and the alternating force F becomes difficult, if not impossible. In order to obtain cancellation ofthe stress-insensitive signals, it is necessaryto ensure thatthe numberofsignals added and subtracted in the memory 12 should be equal. This can be done by counting the numbers of positive and negative multiplications and terminating the process when these numbers become equal.

Figure 2 shows a system which can be used when the applied alternating force varies in an irregular manner. Those components which serve the same functions as similar components in the arrangement already described are given the same reference numerals.

In the embodiment ofthe invention shown in Figure 2, the signals from the strain gauge 6 are digitized and divided in a digitizer/divider 20 into ranges corres ponging to expected ranges of strains in the specimen 1. The digitizer/divider 20 produces signals which are applied to a multi-channel switch 21, to which digitized return signalsfrom the transducer 5 also are applied. The switch 21 actstocausethesignalsfrom the transducer 5to be added into a bank 22 of separate memoriesaccordingtotheir relationship with the bands of strain signals.This addition is continued until a given test run is over and the distribution of strains withinthevarious bands is known.The appropriate additions and subtractions between the memories in the bank of memories 22 is then carried out in the computer 23, and the result displayed on the display unit 13.

Figure 3 illustratesa system which utilisesthe fact thatthe majority ofthe reflection signals are un affected by stress changes, and so give an identical pattern of signals after each transmitted pulse. It is possible therefore to digitize andstore an approximation to this pattern of signarsatthe beginning of a period of measurement, andito subtract at analogue version of this approximationfrom all subsequent echo signal patterns before they are digitizedlThus it is then necessary to digitize onlythe differences between the initial and subsequent reflected signal patterns.

Referring to Figure 3, again, those components which have the same function as ih the embodiment of the invention described with reference to Figure 1',, have the same reference numerals. The additional components area pre-amplifier31 and an analogue subtraction circuit 32 situated between the transdueer 5 and the receiver 7, an additional memory33, and a digitalto analogue converter 34 which is underthe control ofthetimer 14.

In use, an initial measurement ofthe steady pattern of reflected signal pulses is made with no stress applied to the specimen 1 and the additional components inoperative exceptforthe pre-amplifier 31, and a digitizedapproximation to this is stored in the additionaltmemory 33 either in the form of a rectified envelope of the reflected pulse signals, or as afull ultrasoniowaveform. The additional components are then made operative, and the measurements made as before. Atappropriate intervals, determined by the timer 14, the stored signal pattern is released from the memory'33, converted into an analogue form and applied'to the subtraction circuit 32 where it is subtracted from the signals proceedihg from the pre-amplE.rer31 .

The-system operates as follows t Theapproximation to the steady signal pattern is arranged to be digitized in 256 steps (8 binary digits).

Then, ifthe maximum echo signal corresponds, say, to 200 rnitlivoltsatthe inputofthe pre-amplifier31,the corresponding digital step size will be 800 microvolts.

When the analog ue version of thisdig itai approximate steady state echo signal is subtracted from the subsequentanalogue echo signals, the resultant differnoesignal also will bean analoguesignal having thefollowing components: (a) An error which, at maximum will be equal to the approximation digital step sizej, (b) Errors arising from the digital to analogue conversion ofthe approximat steady state echo signal, (c) Errors arising from thesubtraction process"and (d) Real changes in theecho signals.

Let us suppose that eadt ofthese componentsofthe analogue signal from thesubtrnctor 32 may be as large as 800 microvolts, sothattheytotal 2.4 millivolts atthe inputto the receiver7,then if these signals are digitized in another 256 steps by the digitizer8,the step size will be 9 microvolts referred to the receiver 7 input level. It can be shown thatthe step size can be up to threetimes the electronic noise level, soothe arrangement should be satisfactory down to electro nic noise levels of 3 microvolts.

The digitized signalsfrom the receiver 7 are added algebraically, according to the changes in the applied stress in the manner already described. Of the four components of the analogue difference signal listed above, the three errors should be independent of the level ofthe applied stress. As long as the errors do not vary in the shorttime intervals between successive subtractions, they should cancel out, so thatthere will remain onlythosechanges inthe echo signalswhich coincide with changes in the applied stress.

Claims (8)

1. A method of detecting stress sensitive discontinuities in a body, comprising the operations of subjecting the body to an alternating or changing stress, directing pulses of ultrasonic signals into the body, detecting ultrasonic signals returning from the body and producing echo signals related thereto, storing temporarily the echo signals and comparing the stored echo signals corresponding to returning signals originating atone level of the alternating stress with those originating at another level of the alternating stress thereby to eliminate those echo signals which do notvaryin accordance with the applied stress.

2. A method according to claim 1 wherein the first and second levels of stress in the body are both non-zero, and there is included the operations of producing signals related to the level of stress in the bodyduring discretetime intervals, correlating the echo signals with the stress level signals, combining the correlated echo signals with a weighting factor such that the echo signals corresponding to the first level of stress have positive weighted values and those corresponding to the second level of stress have equallyweighted negativevalues, and adding the weighted correlated echo signals thereby to eliminate those echo signals which do notvaryin accordance with the applied stress.

3. A method according to claim 2 wherein there is included the operation of producing echo signals corresponding to a zero level of stress in the body and subtracting these from those corresponding to the first and second levels of stress priortotheir correlation with the stress level signals.

4. A method according to any of claims 1,2 or3 wherein the stress alternates at a known repetition frequency and the pulses of ultra sound are injected into the body at a repetition frequencywhich is an even integral multiple of the repetition frequency of the alternating stress.

5. A method according to claim 1,whereinthe applied stress is irregular and there is included the operations of choosing first and second levels of strain in the body arising from first and second levels of expected range of stress applied to the body, dividing the range between the first and second levels of strain into a number of even intervals, correlating the echo signals with the strain intervals and storing separately the echo signals correlated with each strain interval, multiplying those stored echo signals accumulated in strain intervals above the mean level of strain buy a positive factor, and those accumulated in strain intervals belowthe mean level of strain by an equal negative factor, and comparing the stored echo signals when the numbers of positive weighted echo signals equal the number of negative weighted echo signals.

6. Apparatus for detecting stress sensitive discontinuities in a body subjected to a varying level of stress, comprising means for producing echo signals corresponding to pulses of ultrasound reflected from discontinuities within the body within discrete time intervals, meansfor producing signals relatingtothe varying levels of stress within the body, means for correlating the echo signals in each time interval with the level of stress in the body during that time interval, meansfor storing temporarily the correlated echo signals, and means for comparing stored signals derived from echo signals generated at one level of stress in the body with those generated at a second level of stress in the body thereby to eliminate the component of the echo signals which is not sensitive to the level of stress within the body.

7. A method of detecting stress sensitive discontinuities in a body, comprising the operations of subjecting the body to an alternating or changing stress, directing pulses of ultrasonic signals into the body, detecting ultrasonic signals returning from the body and producing echo signals related thereto, storing temporarilythe echo signals and comparing the stored echo signals corresponding to returning signals originating at one level of the alternating stress with those originating at another level ofthe alternating stress thereby to eliminate those echo signals which do not vary in accordance with the applied stress.

8. Apparatus for detecting stress sensitive discontinuities within a body substantially as hereinbefore described with reference to the accompanying drawings.