GB2048482A

GB2048482A - A method for recognising boundaries or contours of specimens during ultrasonic testing

Info

Publication number: GB2048482A
Application number: GB8013652A
Authority: GB
Original assignee: Krautkraemer GmbH and Co; Krautkraemer GmbH
Current assignee: Krautkraemer GmbH and Co; Krautkraemer GmbH
Priority date: 1979-04-30
Filing date: 1980-04-25
Publication date: 1980-12-10
Also published as: DE2917510A1; BR8002764A; FR2455728A1; IT8048531A0; DE2917510C2; JPS55151208A

Abstract

A specimen 3 in the form of a strip is advanced towards a row of ultrasonic probes 5. The probe signals SK which occur as the specimen boundary intercepts the ultrasonic beams of the probes are evaluated to determine the width of the specimen and other information concerning its contours using equipment programmed to treat interference signals SS as such. <IMAGE>

Description

SPECIFICATION Improvements in ultrasonic testing This invention relates to a method of detecting particular contours of a specimen undergoing ultrasonic testing.

A specimen in the form of a strip or plate may be tested ultrasonically for flaws by transporting it through a testing station including probes for transmitting ultrasonic signals into the workpiece and receiving such signals leaving the workpiece. In accordance with the through transmission technique or principle, for example, a row of transmitting probes is located either above or below the path of the specimen and a row of receiving probes is located on the opposite side of the path. Each receiving probe is in alignment with a respective transmitting probe.Each probe may be housed within a respective probe chamber provided with a valve for discharging water or other coupling liquid for accoustically coupling the probes to the specimen, see for example J and H Krautkrämer, Werkstoffpriifung mit Ultraschall (Testing materials with ultrasonics"), published by springer, third edition, page 407 et seq.

A unit consisting of two aligned probe chambers, one having a transmitting probe and the other having a receiving probe, is referred to herein as a T and R unit or combination. Such a unit scans a single track across the specimen, the width of the track being determined by the dimensions of the probe chambers. As the beam of ultrasonic energy must be narrower than the probe chambers, a 'dead zone' exists between each track and the next. To eliminate the dead zones, or at least to reduce their overall area, additional rows of T and R units may be arranged parallel to the first row but offset relative thereto by an appropriate fraction of the inter-unit pitch the first row.

At the beginning of a test, the T and R units are released, that is to say, the probes become operational electro-acoustically, and the valves are opened to bring about the discharge of the coupling fluid. To reduce the fluid flow rate, those units beyond the lateral or side edges of the specimen may be inhibited, the valves remaining closed and the probes remaining unactivated.

In order to determine the width of the test specimen or to establish its lateral contours and thereby effect the release of the appropriate T and R units, mechanical auxiliary devices, e.g.

pendulum switches, may be used. Thus there may be arranged above the transport conveyor a transverse pendulum which, by means of a lever mechanism, operates electrical switches, see G. Künne, 6. ICNT Report L5, Hannover, 1-5 June, 1970.

These mechanical sensing devices are very disadvantageous, particularly as they have to be as close as possible to the level of the conveyor in order to minimise the length of the pendulum.

This, in turn, calls for a cooling device when transporting hot plates. Transporting of plate or strip-shaped test specimens in the reverse direction on the conveyor is only possible if the pendulum arms are hinged upwards. The entire mechanism for operating the switches, which becomes very dirty when in practical use, is subject to a high rate of wear and a considerable maintenance expenditure. In addition, in the design referred to, this method does not allow contours of the leading and trailing edges of plate or plate-shaped test specimens to be determined with the automatic ultrasonic test and to evaluate them for the test. As a result, the inclusion of this information in a test report is not possible.

It is the object of this invention to provide a method which makes possible the recognition of the width and leading end and lateral contours of strip or plate-shaped test specimens without expenditure on ancillary mechanical devices which require a great deal of maintenance, but which instead takes advantage of the existing electronic equipment such as digital computer units, and which allows release or inhibition of the T and R units in accordance with the contour recognition, while making contour recognition available for data evaluation.

This object is achieved in accordance with the invention by providing a method for automatically recognizing the width, the leading end and side contours of test specimens when testing materials ultrasonically using a plurality of ultrasonic probe units which are arranged in one or more rows and coupled acoustically to a test specimen by means of fluid, preferably water, wherein the intake or discharge of a contour which limits the test specimen in the coupling distance of one or more ultrasonic probe units, there follows a change in the ultrasonic pressure amplitude transferred over the respective coupling distances and this change in amplitude is used as a recognition signal for the passage of a test specimen.contour through a respective coupling distance between the transmitter and the receiver probe chamber and the test specimen, and the changes of ail received ultrasonic pressure amplitude signals result in the total contour and thereby also the position of the test specimen relative to the arrangement of the probe units, these changes in the sound pressure signals being used for the release of the testing channels for indication and/or recording the contour of the test specimen and/or the position of the test specimen and/or the control of further work sequences.

The invention will now be described by way of example with reference to the accompanying drawings, wherein; Figure 1 is a temporal sequence diagram for the contour recognition; Figure 2 is a plan view of a testing installation; Figure 3 is the front view of the testing installation; and Figure 4 is a status table of the electronic memories; Referring to the drawings, a strip or plate-shaped test specimen 3 is transported on a conveyor 7 through a testing station at which is located an ultrasonic testing device 8. When the leading end of the specimen passes a sensor 13, operation of the complete evaluating device is initiated. The sensor 1 3 may respond to a water barrier or light barrier, for example, or the sensor may include a mechanical or other type of switch.The testing device includes a plurality of pairs of probe chambers for an equal number of T and R units 4 arranged in rows 5a, 5b. The probe chambers are associated with flap valves 1 2 or other valves which are opened at the beginning of the test to release jets of water used to acoustically couple the probes and the test specimen, for the purpose of Figs. 1 to 4, there are assumed to be twentyfive units in the row under consideration, the testing channels being designated P, to P25.

The ultrasonic signal generated by the transmitting probe of each T and R unit is received by the receiving probe of that unit. The strength of the received signal will depend first upon whether the unit is operating correctly, secondly upon whether a portion of the specimen lies in the path between the two probes, and thirdly upon whether there is some other occurrence which distorts the transmission. The receiving probes are connected by suitable circuitry with equipment including various memories for evaluating the received signals. Fig. 4 indicates how entries are made in the memories for the example shown in Fig. 1.

Initially, it is necessary to determine whether the testing channels are operating correctly: Testing Channel Check This test is carried out in a time interval to during the period of time tv required for the leading edge 1 of the specimen 3 to travel from the sensor 1 3 to the entry to the probe arrangement 8.

The proper functioning of the T and R units is checked, as is the coupling distance 6.

As will be seen from the table which follows, the correct signal level for each unit, in the absence of the specimen is assumed to be 5a. A monitoring device determines whether, in respect of each channel, there is being received a signal Sa of desired amplitude, and the evaluation is made on the following basis: 1. A signal Sa of preset amplitude consistently present:- testing distance steady.

2. A signal is present but during a short but previously set time duration (e.g. the delay time of a water spray), the signal fails to reach a preset value:- testing distance needs adjusting.

3. A signal is present but its amplitude consistently fails to reach a preset value - the testing distance is defective.

In the event that the last mentioned condition prevails, the T and R units pertaining to the defective tracks fail to deliver signals of acceptable amplitude. In consequence, during the entry recognition phase (q.v.), the defective track is deemed not to be present and the tracks on either side are treated as being adjacent tracks. In Fig. 1, channel 1 3 is so treated.

Following the testing channel check and after the elapse of the period to, the system enters an entry recognition phase: Entry Recognition: This phase occupies the period of time tE during which the specimen travels through the distance 1E The distance 1E depends upon the profile of the leading edge of the specimen and the speed of the conveyor, and the distance LE is predetermined for the given application.

During this phase, the presence in one of the testing channels of a signal (e.g. Sj) of an amplitude different from the steady-state signal Sa results in an entry being made in a memory SPV in a store pertaining to that channel, provided that the signal persists for a predetermined time to prevent the memory being triggered in response to interference. The actual time (tz) at which the signal was received is entered into an appropriate store of a memory SPt.

Thus referring to Figs. 1 and 4, signals are shown to appear on channels P4 to P20 as the leading edge 1 passes through the row of probes during times t, to t,7 in reverse order. In the case of channels P21 and P2, signals appear at times t,7 and t18, respectively following occurrence unconnected with the contours. If, during the entry recognition time period tE, a further signal appears on the same channel, the action taken is as follows: a) if the channel has still not been confirmed for which, see the section entitled "signal confirmation and auxiliary memory"), the former querying time is carried over.

b) if the channel has been confirmed, and released by an auxiliary storage, the new querying time is stored in memory Spz.

The recording of the signal is controlled by a fixed querying cycle in a time slot pattern of a few milliseconds. The established signal times which are also stored in memory Sp2 therefore form a multiple of this time slot pattern, the internal time scale of the computer.

Further processing and evaluation of the signal appearing on any given channel depends upon what is taking place in adjacent channels, and therefore a proximity check is carried out.

Proximity Check: Within the time period tE a pure data recording takes place time-coincidental to it, delayed by the time tT' the proximity check of the individual scanning tracks. In the time period tE, which, as is the case with all the time ranges, is given an electronic time gate, there is a short time gate for the time tT. It corresponds to the time which would be taken by, for example, a quantity of the coupling fluid to splash or spray from one coupling distance to an adjacent coupling distance and there to give rise to an interference signal S,. This time tT is predetermined.

The time gate for tE is started by a predetermined delay, which results from the geometrical factors and the testing speed, whereby release is effected by means of the sensor 13, so that a time gate is already formed, before the leading edge of the specimen has reached the first row of probes 5a. The time gate for tT is started by the gate for tE, which with its duration (tT < tE) passes step-wise, in the time steps of the time slot pattern of the querying cycle through the time gate for tE. The time gate for tE is closed if the proximity check has been concluded. The time duration tE is preferably controlled by the rate of feed of the test specimen.

Each scanning track now begins with the start of the time gates for tE and tT, in order to be able to determine in each adjacent testing channel if, during the time range tT, a signal is also present there. In dependence upon this signal, the signal of the adjacent scanning track is identified either as an interference signal Ss or as a contour signal SK. If the adjacent occurrence is recognized as having come from the contour of the test specimen 3 then a signal confirmation will follow as described in the following section.

Signal Confirmation and Auxiliary Memory: Signal confirmation is carried out by making entries in certain memories in accordance with the results of the proximity check.

Accordingly, confirmation in the memory Sp, takes place for a given channel provided that signals occur within the dead time ti in at least two adjacent channels. This is the case for the channels P4 to P20.

However only one test channel with a confirmed signal for adjacent channel results in an entry in the auxiliary memory Sp,.

In the event that the contour of the leading edge is irregular or wavy several confirmation pairs may be given at the beginning of tE. However during the continued passage of the specimen, the respective adjacent scanning tracks can be confirmed again by channels which have already been confirmed. Also for the other channels the time tT applies as an individual time criteria to the adjacent channel.

The function of the two auxiliary memories Sp, and Sp2 is to recognize large specimen deformations, sprays of water, defective testing channels and other interfering influences and to ensure that in spite of these interfering factors, there is no interference with the determination of the specimen width. If, during a certain time in a confirmed testing channel there is no signal in a neighbouring channel during the dead time tT an entry is made in the auxiliary memory Sup,,.

In the figures, see, for example, the testing channels P 3, P 13 and P 21. If in the other channel (ie. the other neighbouring channel from that which has set Sp1) there is no signal in the dead time tT and for that channel Sp, has already been set, the auxiliary memory Sp2 is set in addition.

In the case of a test channel in respect of which the auxiliary memory Sp, has been set and, which in addition has both sides confirmed testing channels, has only exceeded the dead time tT of one of the two adjacent channels. The signal confirmation is given based on the other adjacent channel during the time tE. However, apart from that there follows an entry in Sp, for the respective test channel. Naturally the auxiliary memories have their own places, as do the other memories, for each test channel.

After the ending of the time range for the entry recognition tE the signal recording with the time gate for tE, as also for the proximity check within the shorter time gate for tT is ended.

Testing channels whose places are set in the auxiliary memories Sp1 and SP2 receive, after the end of the time gate for tE, subsequent signal confirmations, with entries in the memories Sp, and/or in Sp,. Hereby it is immaterial whether an occurrence signal is present or not in memory Sp, for the corresponding test channel.

This subsequent confirmation just described can be carried out, per test specimen, for more that one test channel. If the memory Sp, has been set in two adjacent channels, then similarly there will follow a subsequent confirmation with the entry in memories Sp, and Sp,. The number of these confirmations is however limited according to the specific use.

If the confirmation sequence has been successfully concluded ie. there are not any further confirmation gaps present in memory Sp,, then a check is made to see whether a plate width 1 has been determined in accordance with previously given minimal values. If the size of the minimal plate width has been reached, or, respectively exceeded, then a deduction of the edge track can follow (see the next following section). On the other hand the release of the scanning track Pf only follows according to the minimal plate width as well as a corresponding report.In all cases, in which the signal confirmation does not establish a homogeneous leading and contour of the test specimen, the release of the scanning track follows in accordance with the maximal plate width as well as a subsequent determination of the plate length, by means of the plate contour tracking (see the section so headed). A corresponding test appears on the report.

In accordance with the examples shown in the figures i) If the memory Sp, for each edge group ie. P 4 to P 6 or resp. P 1 8 to P20 has no entry, then the outer channels (here P 4 and P 20) respectively are identified as the immediate edge channels. The edge channels and those channels which are not recognized as edge channels are inhibited for the evaluation of the ultrasonic data, namely the channels P 1 to P 3 and the edge channel P 4 on the one side and on the other side the edge channel P 20 and the channels P 21 to P 25, in the example of Figs. 1 to 4.

ii) If at least one memory place of memory Sp, has been set in one edge group then all channels of both edge groups, as well as those channels which are recognized as being outside the test specimen, are inhibited for the evaluation. Thus, in the example given it is only channels P 7 to P 1 7 which are released for the evaluation and then follows the continuous plate contour tracking via the amplitude evaluation and the through transmission signals.

b) With double-row operation ie. the T and R units are arranged in two rows 5a and 5b and are off-set respectively, by one half of the unit pitch as shown in Fig. 2; Independently of whether, with the confirmation of test channel 1 of test row 5A, the memory has been set or not, the release of the test channel takes place in the 2nd test row 5b from an external pair of probes to the other external pair of probes, that corresponds to test channels P 7 to P 21. The entry identification limits itself for each edge of the test specimen to two edge tracks.In accordance with the input signals either the immediate edge channel of the 1st or 2nd test row and the other one which is further out ie. the one which lies outside the recognized width of the test specimen, or from each row of test specimens, all three test channels are inhibited if the input signals are not plausible. For example input signals are considered as being not plausible if the outer channel delivers a signal but a signal is not delivered by the inner adjoining channel.

Tracking the plate contour After the deduction of the edge track has been completed the evaluating testing channels Pf are established in the test channel memory Spp. The edge contour is monitored on each side of the test specimen by two testing channels PK on each edge. They connect to both sides of the edges of the test specimen, to the test channels Pf which have been released. The actual through-transmission amplitude of test channel PK serves as evaluating criteria which is routed to the evaluating system in a logarithmic scale, namely in linear dB units, in digital or analog form with high resolution accuracy. In practice the amplitude difference, with a previously given test distance, between a free-water jet and one which is covered by a test specimen is 1 8 dB or more.With a resolution of 256 bits per hour 100 dB this corresponds to a digital signal deviation of at least 45 bits.

If an edge countour 2a of this test specimen runs as shown in position K, (K, lies spatially under the row of probes 5a) from the established and released testing track ranges including up to the edge track, first of all P 21, which is outside the plate, an approximately 1 8 dB = 45 bit reduced signal Si is transferred instead of the signal Sa Due to technical drafting reasons in Fig.

1 the test specimen 3 is shown before it enters testing -row 5a and thereby positions K, and K2 are still outside. If this signal deviation is present preferably in at least 2 sequential querying cycles and subsequently the signal for channel P 21 remains, in these querying cycles, at the lower amplitude of the Si signal this means an expansion (contour expansion) of the test specimen on this side. The next channel P 20 is released by means of this contour tracking, P 21 is then the new inhibited edge channel. This information is stored in memory Spp.

If an edge contour 2b of the test specimen, as shown at location K2, runs out of the established and released testing track range including beyond the edge track, and generates thereby a higher signal in edge track P 4, namely Sa, instead of previously Si, once again there will be a signal deviation of approximately 45 bits. If this follows, preferably, in more than 2 sequential querying cycles, if the original lower signal Sa of test channel P 4 on the higher value Sa, thus indicates a reduction in the width (contour constriction) of the test specimen on this side.

By means of this contour tracking the next channel P 5 is inhibited and is evaluated as a new edge channel. This information is stored in memory Spp.

After the deduction of the edge track and with the beginning of the tracking of the plate contour indications, messages appear regarding the width and shape of the test specimen as well as the position on the transporting level, on a printer or plotter and/or on a dialog unit and/or a function tableau. This actual data can, in addition, be used for controlling other devices e.g. for positioning an edge testing installation.

As an alternative to this method it is also possible to carry out contour recognition using the pulse-reflection method. With the pulse reflection method a signal is present in the test channel only if a portion of the test specimen is located in the coupling distance of this channel. This signal can either be the reflection of the ultrasonic pulse or the sound entry surface of the test specimen (entry echo) or that of the backwall (backwall echo).

A comparison signal, unlike in the through-transmission technique shown, is not inevitably present. It can however be obtained whereby a threshold value is formed at a previously determined height e.g. with 1/3 of the modulatability of the receiver amplifier (which can be 35dB) whereby the smallest expected entry - or backwall echo with an uninterrupted coupling modulates the amplifier to 2/3 of its modulation range (ie. 70 dB).

If there is no occurrence of signals over more than two testing cycles, which exceed the 35 dB threshold, no test specimen or part of a test specimen is in the corresponding coupling distance. All signals below this threshold value are then Sa signals.

If a signal occurs over more than two test cycles, which exceeds the 70 dB threshold, then that means that one test specimen, or part of a test specimen, is present in the corresponding coupling distance. These signals which exceed the 70 dB threshold are namely S signals.

Hereby consideration is to be given to the fact that the Sa signal is smaller than the signal S signal.

The invention makes it possible, without mechanical or other auxiliary switching devices, and by only using the scanning tracks and the kind of digital electronics which exist in an automatic installation of this type, to recognize the contour, the incoming shape of the leading end of the specimen and the side contours of plate or plate-shaped test specimens, for the testing operation to evaluate only the testing channels which, with certainty, lie with their coupling distances in the range of the test specimen and, if necessary to take advantage of them for subsequent working sequences.

Reference is made to the specification of our Federal German application No. p 29 1 7 501.8, (a copy of which is on file of the present application) to the extent that the disclosure thereof differs from the foregoing.

For convenience certain of the symbols used in the foregoing are listed below: 1B - width of the test specimen 1E - the distance to be travelled by the test specimen for the entry recognition - - the distance to be travelled by the leading edge of the test specimen from the sensor 1 3 to the probe chamber-row 5 Pj - test channel, i = 1 . . n, if n channels are available.

Pf - released testing channel Sa - signal of a testing channel outside the range of the test specimen S - signal of a test channel within the range of the test specimen SK - contour signal Ss - interference signal Sp, memory for signal confirmation SPN - memory for subsequent signal confirmation Spp - test channel memory Sp, memory for existing signals Spz - memory for signal times Sp, - auxiliary memory 1 Sp2 - auxiliary memory 2 tc - time for checking the testing channels tE - time for entry recognition tT - dead time for checking adjacent tracks.

tv - delay time. Time required to pass through distance lv tz - the point of time of a signal K, - assumed entry point of contour 2a in channel P 21 K2 - assumed exit point of contour 2b from channel P4 CLAIMS

Claims

1. A method for automatically recognizing the width, the leading and the side contours of test specimens when testing materials ultrasonically using a plurality of ultrasonic probe units which are arranged in one or more rows and coupled acoustically to a test specimen by means of fluid, preferably water, wherein the intake of discharge of a contour which limits the test specimen in the coupling distance of one or more ultrasonic probe units, leads to a change in the ultrasonic pressure amplitude transferred over the respective coupling distances and this change in amplitude is used as a recognition signal for the passage of a test specimen contour through a respective coupling distance between the transmitter and the receiver probe or between the probe chamber and the test specimen, and the changes of all received ultrasonic pressure amplitude signals result in the total contour and thereby also the position of the test specimen relative to the arrangement of the probe units, these changes in the sound pressure signals being used for the release of the testing channels for indication and/or recording the contour of the test specimen and/or the position of the test specimen and/or the control of further work sequences.

2. A method in accordance with Claim 1, wherein the through-transmission technique is employed, and the scanning track at each lateral edge of the test specimen with the range of the test specimen, with its through-transmission signals, is used for establishing the lateral limiting contour in such a way that a comparison of the signal amplitude (Sa) follows from the scanning track immediately close to the limitation of the test specimen and first of all outside the area of the test specimen with the signal amplitude (si) from the scanning track close to the limitation of the test specimen and within the range of the test specimen, and if within more than two or another previously given number of testing cycles in sequence, the amplitude difference (sa - si) does not attain a previously given value and the signal amplitude (sa) similarly does not attain another previously given value, this is indicated or evaluated as an increase in the width of the test specimen on this side of the test specimen.

3. A method in accordance with Claim 1 wherein the through transmission technique is employed, and the scanning track present at each lateral edge of the test specimen within the range of the test specimen, with its through-transmission signals is used to establish the lateral limiting contour in such a manner that a comparison of the signal amplitudes (sa) follows from the scanning track immediately close to the limitation of the test specimen and outside the range of the test specimen, with the signal amplitude (si) from the scanning track close to the limitation of the test specimen and, first of all, within the range of the test and, if within more than two or another previously given number of sequential testing cycles the amplitude difference (sa - si) does not attain a previously given value and the signal amplitude (si) itself exceeds another previous value, this is indicated or evaluated, as a reduction in the width of the test specimen on this side of the test specimen

4. A method in accordance with Claim 1 for differentiating between contour and interference indications which, for example, could occur by a splashing of the coupling fluid, wherein a first time gate is present and recognition signals, which occur within this first time gate are only evaluated as contour signals if the ultrasonic probe units, which are arranged in rows, recognize a signal for each neighbouring track, which lies within a second gate, whose opening time is less or equal to a possible dead time or the neighbourhood control and that cyclicly, the cycling is done through the first gate.

5. A method in accordance with Claim 1, wherein upon the energy of a contour that limits the test specimen in the coupling distance of a receiver probe for pulse-reflection operation, a reflection signal occurs from a surface of the test specimen and that this signal is used as a recognition signal for the entry of a test specimen contour in this coupling distance, and that the occurrence of this signal, in the corresponding coupling distance results in the complete contour and thereby the position or change in the position of the rest specimen on conveyor with regard to the arrangement of the probe units, and that these recognition signals are used for indicating and for recording the contour of the test specimen and/or for controlling other work sequences.

6. A method in accordance with Claim 1, wherein when a contour which limits the test specimen leaves the coupling distance of a transmitter-receiver probe during operation by the pulse-reflection technique, an existing signal drops to a previously established difference value, and the non-attaining of this difference value is used as a recognition signal for the discharge of a test specimen contour from this coupling distance and the occurrence of this non-attaining of these difference values results, in the corresponding coupling distances, in the complete contour and thereby the position of the change in the position of the test specimen on the conveyor with regard to the arrangement of the probes and that these recognition signals are used to indicate and/or record the contour of the test specimen and/or the position of the test specimen and/or for controlling further work sequences.

7. A method of ultrasonic testing substantially as hereinbefore described with reference to the accompanying drawings.