NO136119B - - Google Patents

Description

The invention relates to a method for ultrasonically measuring the thickness of an elongated element where relative motion exists between successive sections of the elongated element and a source for electric ultrasonic pulses, and first and second echo measurement pulses from. different, opposite surfaces of the element are distinguished from noise and unused echo pulses by a qualifying procedure comprising the steps of sending an ultrasound pulse into the element, receiving the first echo pulse of a minimum amplitude in response to the ultrasound pulse from one surface of the element, receiving the associated second echo pulse with a minimum amplitude from an opposite surface of the element within a given time period, and to develop signals after receiving the first and second echo pulses with a predetermined minimum amplitude

them, the time duration between the signals being related to the elapsed time between the reception of the first echo pulse from one surface of the element and the reception of the corresponding second echo pulse from the opposite surface of the element. Furthermore, the invention relates to a device for carrying out the method.

The invention deals in particular with ultrasonic measurement of the thickness and eccentricity of a sheath which is extruded around successive sections of a cable which is fed forward along a predetermined path.

In an example of a cable designed for use in the communications industry, a metallic moisture barrier is wrapped longitudinally around a cable core which is fed along a predetermined path with the overlapping edges of the moisture barrier stapled together to form a watertight seam. A sheath of plastic material is then extruded over the core and moisture barrier with the sheath in certain cable types bonded to the outward facing surface of the moisture barrier. The sheathed cable is then passed through a relatively

long water chute to reduce the jacket temperature.

It is desirable to continuously monitor both the casing wall thickness and the casing's eccentricity. Monitoring of the jacket facilitates maintenance of a uniform thickness of the jacket and saving plastic material by applying the minimum acceptable jacket thickness. The monitoring of the eccentricity is helpful when making adjustments to the extrusion apparatus to prevent the finished product from being "out of round".

The previously known devices control the thickness of a plastic sheath which is extruded over a metallic moisture barrier by measuring the capacitance between the metal barrier and an electrode which is placed in direct contact with the plastic sheath. As there is no metallic moisture barrier surrounding an inner jacket, conventional dynamic capacitance methods cannot be used and "off-line" tests must be used.

For effective control, it is desirable to check the sheath thickness as soon as possible after the plastic has been extruded onto the cable core. With the previously known devices, the check is carried out after the cable has been led out of the water cooling channel in order to avoid connecting the electrode together with the cable jacket until the temperature of the cable jacket has been reduced. A non-contact shell thickness measuring device near the point where shell extrusion is performed essentially optimizes the feedback control system coupled to the extruder or provides results that are displayed for use by an operator when making adjustments to control eccentricity and shell thickness . This will allow saving of cladding material by continuous control of its placement.

Previously known measurement systems are shown and described in

the US patents 3,407,352 and 3,500,185 which show magnetic coil sensors and capacitance pickup sensors feeding bridge circuits respectively.

Ultrasonic methods for measuring the thicknesses of solid bodies have been used in the prior art and are usually centered on two methods. One is a pulse echo method where the pulse is transmitted and the system waits for the echo to return.

Another ultrasonic method is what is referred to as a resonance method. This technique works according to the principle of an increase of a return signal when the wavelength of the ultrasound signal approaches an integral multiple of the thickness of the object being tested,

The application of the ultrasonic pulse echo technique for measuring thickness and eccentricity seems to be ideal. A crystal is stimulated or shock-excited electrically by applying a high voltage to form a mechanical self-oscillation in the crystal. The mechanical movement is transmitted in the form of a pressure or sound wave through the coupling medium, e.g. the water in the cooling trough in the present description, to the cable being measured.

Depending on the acoustic mismatch between the coupling medium and the article being tested, e.g. a cable, some of the sound wave energy will be reflected and some will continue through the article being tested. A similar transformation occurs at each interface in the article. The echo signals are processed to produce a measure of the thickness of the mantle.

In the case of ultrasonic measurement of elongated materials, such as pipelines, in the past a crystal with a line focal point has usually been placed with the line focal point parallel to the longitudinal axis of the pipeline. In pipe fabrication, however, the rigidity of the pipe precludes any significant lateral movement of the successive sections of the pipe. This is different from the lateral movement that occurs when the successive sections of a cable are moved through the cooling channel.

Previous patents showing ultrasonic measurement methods include US patents 3,423,992, 3,474,664, 3,509,752 and 3,605,504.

However, the prior art does not contain aids which are adapted to the special needs of checking the thickness and eccentricity of successive sections of a cable sheath immediately after extrusion, and which are able to measure only those signals which have meaning as the cable is fed forward. The purpose of the invention is to provide a method and a device for efficient performance of such control.

The initially stated method is, according to the invention, characterized by the fact that the first echo pulse is confirmed by testing with regard to a first predetermined timing characteristic comprising duration in addition to - minimum amplitude, and the second echo pulse is confirmed by testing with regard to a second predetermined timing characteristic comprising duration in in addition to an amplitude that is greater than the minimum amplitude, since the signals that are developed upon reception of the first and second pulses are respectively distinguishable on and off signals, the on and off signals being developed only in response to the reception of first and second echo pulses having at least the minimum amplitude, and the off pulse being developed only if a predetermined time interval has passed after the beginning of the first echo pulse.

A device for carrying out the stated method comprises a device for sending an ultrasound pulse to the element and then receiving a first echo pulse that is reflected from one surface of the element and then receiving an associated second echo pulse from the opposite surface of the element, and a device for qualifying the first echo pulse and then the associated second echo pulse based on amplitude, and is characterized in that it comprises a device for, in response to the receipt of a qualifying first echo pulse, to initiate development of an on signal and examination of a confirmatory test characteristics of the first echo pulse, means for, in response to the confirmation of the first echo pulse and upon the receipt of a qualifying associated second echo pulse, to initiate examination of confirmatory test characteristics of the qualifying second echo pulse, means for cause the development of an off signal where the duration between on -signal and the off-signal are related to the elapsed time between the reception of the first echo pulse from one surface and the reception of the associated second echo pulse from the opposite surface, a device that measures the time duration between the on and off signal, a device for , partially in response to the confirmation of the first echo pulse, preparing means for enabling the generation of a control pulse, a device for completing the preparation of the enabling means in response to the confirmation of the associated second echo pulse to generate the control pulse,

and means for processing the measured duration in response to the generation of the control pulse.

More specifically, with the method according to the invention, a test period is established during which echo pulses reflected from a cable sheath exposed to ultrasound waves are tested. An ultrasound pulse is introduced into the cable jacket and then a first echo pulse that is reflected from one surface of the element, and an associated second echo pulse that is reflected from an opposite surface of the element, are received, while essentially isolating the reception of extraneous low-level signals.

The first echo pulse is sampled, and then an associated second echo pulse is sampled with respect to a qualifying sample characteristic or sample property. In response to receipt of a qualifying first echo pulse, the development of the on signal and the examination of a confirmatory sample characteristic of the first echo pulse are initiated, thus, in part in response to the confirmation of the first echo pulse, to prepare means to enable the generation of a control pulse. In response to the confirmation of the first echo pulse and to the receipt of a qualifying, associated second echo pulse, the examination of confirmatory sample characteristics for a qualifying, associated second echo pulse is started, while causing development of the off signal. The time duration between the on signal and the off signal is related to the elapsed time between the reception of the first echo pulse and the reception of the associated second echo pulse. The elapsed time between the on signal and the off signal is measured and the preparation of the enabling means is completed in response to the confirmation of the associated second echo pulse, to generate the control pulse. The measured elapsed time is processed in response to the generation of the control pulse to provide an indication of the sheath thickness.

The invention shall be described in more detail below

by means of a number of design examples with reference to the drawings, where fig. 1 is an overview view in perspective of an equipment for extruding a jacket of plastic material around successive sections of a core of insulated conductors, and where there is shown an ultrasonic equipment that uses the present invention for measuring the thickness and eccentricity of successive sections of the cable jacket, fig. 2 is an enlarged, partially sectional view of a water trough or water trough indicated in FIG. 1 and which shows the placement of four converters

in the form of crystals relative to the jacketed core fed forward through the trough, and also showing a side view of a console for displaying the measurement in analog form and for displaying the eccentricity of the cable sheath, Fig. 3 is a schematic overview showing a typical one of four channels used to perform ultrasound measurements, and which shows components that are common to all four channels, fig. 4 is a connection core of a pulser receiver unit shown in FIG. 3 for exciting one of the converters to emit pulses to the cable, and for amplifying returning pulses, fig. 5 is a schematic diagram of a receiver logic circuit shown in FIG. 3 for discriminating between false signals and meaningful measurement signals, for controlling the generation of a read pulse for confirmation of a time interval indicating the thickness of the jacket, and fig. 6 is a diagram showing waveforms associated with the operation of the devices in FIG. 3 showed equipment, and which have suitable inscriptions for indicating the different curves in the diagram.

In fig. 1 shows a device 10 for coating successive sections of a cable core 11 with a plastic material, such as polyethylene, to provide a cable 12 with a jacket or sheath (see Fig. 2). As shown in fig. 1, the equipment 10 comprises an extruder 14 and a cooling trough 16 which is placed close to the extruder on its downstream side. Successive sections of the core 11 are fed by means of a walking winch 17 forward through the extruder 14 where the plastic material is extruded onto the core. Next, successive sections of the cable 12 are fed into and through the cooling trough 16 and wound onto a spool 18.

In order to most effectively control the thickness "d" of the jacket 13 (see Fig. 2) which is extruded over the core 11, and also to control the eccentricity of the jacket, an ultrasonic jacket measuring device generally designated by the reference numeral 20 (see Fig. 1 and 2), arranged in the equipment 10. The device 20 monitors or controls the thickness and eccentricity of the jacket immediately when the successive sections of the jacketed core 11 are fed into the water trough 16. The ultrasonic jacket thickness measurement device 20 is designed to use the well-known pulse echo measurement technique which requires at least one crystal 21 (see fig. 2). The water in the cooling trough 16 serves as a coupling medium for transferring ultrasonic energy to the cable jacket 13.

To carry out a meaningful control of the cape

13 thickness d, it is necessary to measure the sheath thickness at several points around its circumference. This is also necessary to be able to determine ©; the eccentricity of the cable jacket 13.

As best shown in fig. 2, a number of crystals is 21-

21 scattered around and separated from the circumference of the sheathed cable 12, each of the crystals being immersed in the coolant, e.g. water, in the cooling trough 16. The crystals 21-21 must be chosen so that they concentrate as much energy as possible in a small area, i.e. it is desirable that the crystals have a line focal point where the line lies across the cable 12. The choice must also be so that the crystals 21 can be used for all expected cable sizes at a manufacturing site, and also for as many different materials as possible.

Another critical property is the crystal's 21 attenuation factor. If the damping is not fast enough, no meaningful indication of the thickness of the jacket 13 will be obtained, i.e. the pulse echoes (echo I and echo II in Fig. 6a) merge and become imperceptible.

Tests were carried out on several different types of crystals. A lead zirconate crystal was found to be most desirable as it produces a stronger signal at the same excitation than what is produced with a 10-MHz lithium sulfate crystal 1., and also with better attenuation characteristics.

The crystals 21-21 are excited by the application of a voltage pulse which causes cyclic mechanical stresses or stresses. These stresses generate high-frequency pressure gradients or waves in the coupling medium, which in this case is the water in the cooling trough. The waves, which vary in intensity as an attenuated sine wave, propagate to the cable sheath's .13 surface (during a time T, see Fig. 6a)) .where ;a -part .of the waves is reflected due to the acoustic impedance mismatch. This causes an echo to be produced from the outer surface (echo I-, see fig. 6a).

Part of the pressure wave also travels. inwards into the cable jacket 13 whereby a second reflection occurs at the inward facing surface of the cable jacket. This determines the second echo or inner surface echo (see echo II, Fig. 6a).

The term "facing" as used in the present description when referring to the relationship between a surface and an object is to be understood to mean that the surface is oriented in the direction of the object, and may be, but is not necessarily, close up to or in contact with the object.

It will be remembered that the measurement technique used is to determine the time separation (2t, see fig. 6a) between echo signals. This separation is directly proportional to the thickness d of the cable jacket 13. The thickness data depends on the uniformity of the speed of sound in polyethylene - Experiments have shown that this speed is essentially uniform, thus allowing the application of the equation 2d = vt where d = the polyethylene thickness, v = the speed of sound and t = the echo separation time. The echo separation time is measured, and since v is known (approx. 0.051 cm per microsecond), d can be easily determined. There may be some unevenness in the rate of propagation at the surface of the polyethylene sheath which cools before the interior of the sheath 13.

Each of the crystals 21-21 is connected electrically to an associated one of a number of channels 22-22 (see Fig. 3). Each of the channels 22-22 comprises a number of elements for transforming the time interval between pulse echoes that are received from the various interfaces of the materials that are aligned in relation to the associated crystal 21, into an output signal that is proportional to the thickness of the cable jacket 13.

There are further arranged other elements which together with the four kahale.r 22-22 comprise an electrical circuit which is generally denoted by the reference number 25. The output signal from the circuit 25 expressed as thickness and eccentricity is presented to an operator on a console 26 (fig. 2).

In fig. 3 shows a typical one of the channels 22-22 together with other elements that are common to the four channels. Each of the channels 22-22 contains the associated one of the crystals 21-21 which is connected electrically to an associated pulser/receiver unit 27. The pulser/receiver unit 27 is a specially constructed device which at the beginning of each cycle is used to send a pulse to the associated of the crystals 21-21 to cause the associated crystal to oscillate at its natural frequency. Then the pulser/receiver unit 27 operates in each of the cycles so that it receives pulse echoes from the cable 12 under test.

The pulser/receiver unit 27 is electrically connected via a conductor 28 to a receiver logic circuit 29 which, on command from a pulse repetition frequency circuit 30, causes the pulser/receiver unit 27 to supply a pulse to the associated converter crystal 21. As shown in fig. 3, the pulse repetition frequency circuit 30 is connected via a conductor 31 to the receiver logic circuit 29. The latter circuit is designed to confirm echo pulses received from the cable 12 under test, and is capable of rejecting those pulses which do not concern the measurement of the sheath 13 thickness . As such, the receiver logic circuit 29 must have inherent intelligence to distinguish between noise peaks and valid signals, and also to differentiate between the first and second echo signals.

The receiver logic circuit 29 is electrically connected to a counter 32 which contains a number of decade counters 33-33. The counter 32 is designed to measure the width of the output pulse from the receiver logic circuit 29. The counter 32 is pulsed using a commercially available oscillator 34 (see the oscillator output signal, Fig. 6i), and is designed to store the pulse count in a buffer or memory store (not shown ) provided that the pulses received by the pulser/receiver unit 27 are confirmed by the receiver logic circuit 29.

The stored digital count from the counter 32 is then transferred to a digital-to-analogue converter 36 which converts this digital count into an analogue voltage. The function of the digital-analogue converter 36 is to provide an analogue voltage which corresponds to the digital count stored in the counter 32. This voltage is an indication of the thickness for the associated channel. The continuous analog voltage is displayed in a suitable scale on a meter 37 (see Fig. 2) which is connected to the relevant channel 22. This allows an operator to continuously monitor the thickness d of the cable sheath at part of the circumference which is linked to the aforementioned channel.

In an alternative embodiment, the counter output signal stored in the buffer can be connected to a commercially available universal digital calculator (not shown) so that data analysis and reduction can be performed.

In order to measure the eccentricity of the cable jacket 13, an eccentricity measurement circuit 38 is connected to the digital-analog converter 36 for each of the channels 22-22. The eccentricity measurement circuit 38 is designed to compare the sheath thickness d at the top and bottom of the cable sheath 13 as shown in fig. 2, and the sheath thickness at the left and right side of this. It is of course not necessary for the comparative measurements to be taken along horizontal and vertical axes, but instead only at opposite intersection points for the axes of a coordinate system with the cable jacket 13.

The eccentricity measurement circuit 38 is designed to subtract the thickness for the bottom measurement from the measurement at the top of the jacket 13 as seen in FIG. 2, the result being multiplied by 100 and divided by the nominal shell thickness to give a percentage of nominal shell thickness. A similar calculation is carried out with regard to the thickness at the left and right parts of the cable sheath 13. Each of these measurements is displayed on a top-bottom gauge 39 and a left-right gauge 40 which are connected to the eccentricity measuring circuit 38.

The pulse repetition frequency circuit 30 is designed to supply a signal (-y_) at a time designated as the PRF time, over a conductor 41 to a connection point 43 for, for example, the first channel and from there along a conductor 44 to an input 47 of a pulse device or optical isolator 48 (see Fig. SX. The pulse device 48 supplies a pulse to the pulser/receiver unit 27 to control this unit so that it causes the associated converter 21 to emit pressure waves.

The pulser/receiver unit 27 contains a trigger circuit

51 (fig. 4). to which the pulse from the pulse device 48 is supplied to cause the supply of a high current pulse to a control electrode 53 in a first controlled rectifier (SCR) 54 to cause ignition of this first controlled rectifier.

The controlled rectifier 54 is a commercially available device and can be a two-state semiconductor device with thyratron-like operating characteristics. The controlled rectifier 54, which is normally in a non-conductive state, comprises a cathode 56 and an anode 57, the latter being biased in the forward direction by means of a suitable external device, such as a positive voltage source 58. When the control electrode 53 becomes suitable energized by a bias potential, it brings the controlled rectifier into conducting state.

As shown in fig. 4, the anode 57 of the controlled rectifier 54 is above a connection point 59 and a connection point 61 connected in series with a cathode 62 in a second controlled rectifier 63. The connection point 61 is connected to a control electrode 64 of the second controlled rectifier 63. The controlled rectifier 63 anode 66 is connected via connection points 67 and 68 to cathode 69 in a controlled rectifier 7.1. The connection point 68 is connected to a control electrode 72 in the controlled rectifier 71. Finally, the anode 73 of the controlled rectifier 71 is connected via connection points 74 and 76 to the cathode 77 in a controlled rectifier 78, the connection point 76 being connected to a control electrode 79 in the rectifier 78.

As further shown in fig. 4, it is controlled rectifier

78 anode 81 over a connection point 82 connected to ground 83.

Connected to the controlled rectifiers 54, 63, 71 and 78 are respective resistors 84, 86, 87 and 88 which are connected in parallel with the rectifiers. The connection point 82 is connected via a conductor 89 to a connection point 91 with which the series of resistors 84 and 86 - 88 are connected, and then across a capacitor 92 to a connection point 93. The connection point 93 is connected to the cathode 94 of a diode 96 whose anode is connected via a connection point 98 to the associated one of the converters 21-21. The diode 96 is effective for blocking spurious noise signals that are of low amplitude. If the diode 96 were not present, low-amplitude noise signals from the ignition part of the circuit would tend to propagate from connection point 93 to connection point 98 and attempt to degrade echo signals from the associated converter 21. In response to these low-level noise signals, diode 96 as an open circuit for either positive or negative output.

The blocking diode 96 is further across the connection point 98, a conductor 101 and a resistor; 102 connected to a connection point 103 and from there to an input 104 of a broadband video amplifier 106. In order to avoid the occurrence of a large negative voltage at the input 104 of the amplifier 106, which would be harmful to the amplifier, a diode 107 is inserted in a conductor 108 which is connected to the connection point 103.

The commercially available broadband low-noise video amplifier 106 receives, amplifies and then outputs return echo signals from the cable 12. The low current return signals must be amplified to a level of approx. one to three volts. The broadband video amplifier 106 is chosen so that the semiconductor devices in it have low noise properties. In this way, the signals received and amplified for transmission to the receiver logic circuit 29 will be stronger than the signals caused by noise, and will therefore allow accurate detection of the meaningful signals.

The provision of aids in the pulser/receiver circuit 27 to reduce the noise level during a period of time when

echo pulses are accepted for testing the circuit 25, is particularly important when the device 20 is used for measuring the thickness of an inner jacket. The pulse echoes received from the inner sheath have a lower amplitude than those received from an outer sheath. This is caused by the fact that the polyethylene surface is more irregular on the inner covers. The weak crests and troughs of the inner sheath reduce the target area or strike surface of the cable, thereby causing less leading edge amplitude for the echo signals. The weak peaks of wave troughs also cause phase differences in parts of the electrical echo signals, which can contribute to a phase shift that gives lower amplitudes. Although the echo signal strength varies between inner and outer sheaths and between different cable sizes, a system with fixed amplification can be used for all types of polyethylene cables.

The inner sheath shrinks around the cable core 11 and gives

a somewhat irregular outward facing surface. It. the outer sheath, on the other hand, is extruded over a tubular metal screen which helps to cause the outward facing surface of the outer sheath to be relatively uniform. This necessitates a setting for a lower threshold interception, which in turn increases the chances of interception

ning of noise signals.

The diode 96 which is interposed between the pulse generator portion of the circuit comprising the controlled rectifiers 54, 63, 71 and 78 and the broadband video amplifier 106 in the pulser/receiver unit 27 acts to raise the impedance level during the time that the receiver 27 is "listening" for echoes. This allows better operation of the wideband video amplifier with respect to the impact of noise on the operation of the device 20.

The diode 96 is inserted into a conductor 89 which connects the pulse generator or pulse shaping portion of the circuit 27 to the wideband video amplifier 106, so that the circuit in a first direction appears to be 'open'. In the other direction, towards the left in fig. 4, the diode 96 allows the passage of a negative current which pulses the associated one of the converters 21-21.

The detection system is adapted to the characteristics or properties of the expected signals. For example, due to the known distance between the transducer 21 and the cable 12, as well as the propagation speed of sound waves in a water medium, the time interval during which pulse echo signals can be expected is generally known. The ultrasound device 20 is therefore designed so that it accepts pulse echo signals only during this time interval, which will be referred to below as the "window width".

Two sound properties of polyethylene, namely the propagation speed and the acoustic impedance, are particularly important. Both are dependent on the polyethylene temperature and therefore require extensive on-line investigation.

The equation for the velocity of the longitudinal wave (displacement wave not important here) in a medium is given as:

where = longitudinal velocity, u = shear modulus, p = density and Kc = bulk modulus of elasticity.

The acoustic impedance of polyethylene is important as it affects the size of the echo signals. The equation for acoustic ixr.-pedance is as follows:

with the same symbols as above. The degree of reflection at the interface between two media increases as the difference in acoustic impedance increases. This degree of impedance mismatch determines the echo amplitudes. The interface between water and polyethylene determines the first echo, and the interface between polyethylene and air, steel, aluminum or flooding compound determines the second.

Experimental data showed that acoustic impedance for polyethylene increased with increasing temperatures. In particular, the first echo had a much lower amplitude than the second echo due to the greater acoustic mismatch at the second polyethylene surface. Furthermore, the second echo signal had a much longer duration than the first echo signal. These signal characteristics are used with advantage in the construction of the "intelligence" in the receiver logic circuit 29. In case of inner casing measurements where the signal amplitudes would vary considerably due to surface irregularities, this "intelligence" is particularly advantageous or useful. If the echo amplitude or echo duration momentarily drops below a certain minimum, no erroneous measurement is made. Instead, the last previous good measurement was held in a register which forms a buffer for the counter 32 to maintain a proper thickness output signal.

The receiver logic circuit 29 is designed to control the following expected signal characteristics or properties: (1) the signal echoes occur in a predetermined time period (referred to as the window width) after the transmitted pulse, (2) the starting polarity of each echo is undetermined, (3) the first echo pulse has a short duration compared to the second echo (see fig. 6a), (4), a time interval of at least 200 nsec. exists between the first and second echo, (5) the amplitude of the second echo pulse is greater than the minimum required for the first echo (see Fig. 6a, and (6) the second echo pulse has a duration of at least 0.700 ysec. If these properties are met for a particular sequence of associated first and second echo pulses, the receiver logic circuit 29 issues an order in the form of a read pulse to the counter 32 to store the count recorded in the time 2t (see Fig. 6h) as valid data.

Connection point 43 is connected to an input 111 of a monostable, single-stage multivibrator 112 (see Fig. 5) which generates a delay pulse which is supplied over a conductor 113 to a second monostable multivibrator 114, and from an output 115. An output 116 of the multivibrator 114 is over a conductor 117 and a connection point 118 connected to an input 119 to a threshold detector or comparator 12 o for detecting a positive voltage, and an input 121 to a threshold detector or comparator 122 for detecting a negative voltage.

The threshold detectors 120 and 122 are included in a comparator part 123 of the receiver logic circuit 29. As shown in fig. 5, the comparator part 123 comprises a low frequency input terminal 124 which is connected to the broadband video amplifier 106. The input 124 is connected to the inputs 126 and 127 of respective threshold detectors 120 and 122.

Detectors 120 and 122 have a negative, true output when their threshold levels are exceeded. As shown in fig. 5, an output 128 from the threshold detector 120 is connected to an input 129 to a NOR gate 131 with negative logic. This terminology is interpreted so that if either the input 129 or another input 133 is supplied with a negative or low level, an output 134 from the NOR gate 131 will have a positive or high level. An output 132 from the detector 122 is connected to the second input 133 of the NOR gate 131. The output 134 of the NOR gate 131 is connected via connection points 136 and 137 to an inverter 138 which changes the polarity of the input signal, and then via a conductor 139 with a flip-flop 141 for setting the flip-flop.

The flip-flop 141 has a reset input 142 connected to the pulse repetition frequency circuit 30, and has an output 143 which, via a connection point 144, is connected to an input 146 to a "hold-off" pulse generator 147. The pulse generator 147, which is a monostable single-phase multivibrator, is over a conductor 148 connected to a confirmation pulse generator 149. This pulse generator is also a monostable single-phase multivibrator.

An output 151 from the multivibrator 149 is connected via a conductor 152 to an input 153 to a positive, true NAND gate 154. As can also be seen from fig. 5, the connection point 136 is connected via a conductor 156 to the second input 157 of the NAND gate 154. If positive signals are simultaneously supplied to the inputs 153 and 157 of the NAND gate 154, a negative or low level appears at its output 158.

The output 158 of the NAND gate 154 is connected via a conductor 159 to the setting input 161 of a flip-flop 162. The flip-flop 162 is reset by means of a delayed PRF signal. This is caused by applying a reset signal to the flip-flop's input 163 using the trailing edge of the pre-sink pulse from the output 115 of the multivibrator 112.

As further shown in fig. 5, an output 164 of the flip-flop 162 is connected via a conductor 166 and a connection point 167 to an input 168, designated as the "D" input, to a flip-flop 169. A clock input 171, designated as the "C"- the input to the flip-flop 169 is connected via a conductor 173 to an output 172 from the multivibrator 149.

The flip-flop 169 which is reset by a pulse repetition frequency (PRF) signal applied to an input 170 produces a signal at an output 174 and this signal is transferred over a conductor 176 to an input 177 of a positive true NAND gate 178. The connection point 137 is over a conductor 179 and a connection point 181 connected to a second input 182 of the NAND gate 178. An output 183 of the NAND gate 178 is over a conductor 184 connected to an input 186 of a flip-flop 187 .

The flip-flop 187 has a reset input 188 which is connected to the pulse repetition frequency circuit 30, but the reset signal is delayed until the trailing edge of the pulse from the multivibrator 112. An output 189 of the flip-flop 187 is connected via a conductor 191 to an input 192 to a positive, true NAND gate 193. The connection point 144 is further connected via a conductor 194 to the second input 196 of the NAND gate 193, and the NAND gate produces a negative output pulse or a low level output pulse as shown in FIG. 5.

Another output 197 of the flip-flop 187 is connected via a conductor 198 to an input 199 of a monostable single-pass multivibrator 201. The multivibrator 201 acts as a generator of a "hold-off" pulse for testing the characteristics of the second pulse echo. An output 202 from this multivibrator is over a conductor 203, a connection point 204 and a conductor 206 connected to an input 207 to a pulse generator 208 for confirmation of the second echo. The generator 208 is also a monostable single sweep multivibrator.

An output 209 of the multivibrator 208 is connected via a conductor 211 to an input 212 of a positive, true NAND gate 213 where a second input 214 of this NAND gate is connected back to the connection point 181 via a conductor 216. The NAND gate 213 is adapted to emit a negative signal on an output 217 and then across a conductor 218 to set a flip-flop 219 across an input 221 to this.

The flip-flop 219 is a so-called ring flip-flop and has a reset input 222 connected to the pulse repetition frequency circuit 30. The flip-flop 219 is further adapted to emit a signal on an output 223 and then via a conductor 224 to an input 226 to a positive, true NAND gate 227. When the NAND gate 227 has positive or high signals applied to both of its inputs, namely the input 226 and an input 228 which is connected via a conductor 229 to the connection point 167, the NAND gate emits a negative or low signal over a conductor 231 to an input 232 of a pulse generator in the form of a monostable single-stage multivibrator 233 to start the multivibrator.

The connection point 204 is connected via a conductor 236 to an input 237 to an amplitude comparator 238 in the detector circuit 123. An output 239 from the comparator 238 is via a conductor 241 connected to a flip-flop 242. The flip-flop 242 also has a reset input 240 which is connected to the pulse repetition frequency circuit 30. An output 243 of the flip-flop 242 is connected via a conductor 244 to an input 246 of the multivibrator 233. Another input 247 of the multivibrator 233 is via a conductor 248 connected back to the output 249 of the multivibrator 114.

The multivibrator 233 is adapted so that on command at the end of the window width it generates a read pulse which causes the count to be stored in memory stores (not shown).

It is important to note that there are approx. 3.9 to 7.6 cm of water above the top of the cable 12 for the device 20 to function correctly. Furthermore, a bottom part of each of the crystal converters 21-21 must be submerged. There may be some difficulty when it comes to fulfilling this requirement with existing equipment.

A way out may be to use a cylindrical device which can be placed above the top of the converters and in close contact with the cable 12. A vacuum can then be applied to the container to draw water up and into it to submerge the bottom part of the enclosed cryst.

One should be aware that the expression "measure" or "test" or "monitor" as used here should be interpreted so that one means comparison of a size with a reference size. For example, the threshold detectors 120 and 122 perform checks to determine whether the size of the echo pulses is at least equal to the ter Threshold amplitude. On the other hand, the comparator 238 determines whether the peak amplitude of the second echo pulse sr is greater than a predetermined amount. The device 20 could of course be designed so that the real values of the time durations and amplitudes are determined.

Depending on the high frequency that can be measured during the monitoring or control process, the device 20 can be provided with minor finesse and still lie within the scope of the invention..- For example, the channel 22 can be designed so that it automatically controls thresholds for both pulses and peak amplitude for the second echo pulse, testing of an arbitrarily recorded second pulse is permitted after a preset time. This of course assumes that the first echo pulse that has the required threshold value is a valid pulse, which may or may not be correct. As an alternative, the device 20 can be designed to determine whether the threshold values for both pulses are above a predetermined value, if the peak amplitude of the second echo pulse is greater than a predetermined value, and if the duration of the first echo pulse is less than a predetermined value.

The preferred embodiment described here registers a time count between the echo pulses if a sequence of first and second echo pulses is confirmed. The first pulse is confirmed if the amplitude is above a certain threshold value and the duration is less than a preset value. The second pulse is confirmed if its peak amplitude is above a preset sturgeon travel and the duration is greater than a preset duration. A valid sequence of a first and second echo pulse must occur in the so-called window width.

The present invention also implies that it is not only the real values of pulse amplitude and duration that can be determined, but also the frequency content of the echo pulses for comparison with those associated with special materials for the jacket being tested. The various subtleties of the present invention are necessary as it is intended for use for dynamic measurement of cable sheath thickness while the cable is fed forward. In contrast to the measurement of pipe wall thickness, the measurement process must take into account lateral movement of the cable and deformations in the sheath 13. With some previously known devices, manual settings must be made when the cable size or other parameters are changed. A device that uses the principles according to the invention adjusts itself automatically after these changes. A logic system must have inherent intelligence to take into account the relatively greater number of imperfections in the cable jacket than that which occurs in the pipeline area.

All the logic elements, i.e. the NAND gates and NOR gates, the bistable elements, i.e. the flip-flops, and the mono-stable elements, i.e. the single-stage multivibrators, are described in operation in connection with positive logic where a voltage with a high or positive level represents a binary "1", and a low or zero level voltage represents a binary "0". Furthermore, throughout the generation method, the input signals and output signals for the various elements are described in connection with high or low levels. It will be understood that this is to be interpreted in accordance with the preceding description.

A positive true NAND gate produces a low-level output signal only when a high-level input signal is present on all of its inputs. The output signal is high level for any other combination of input signals.

A negative true NOR gate produces a high level output signal when a low level is present on a positive input. Only .when there is no low-level input, the output is low-level.

A flip-flop is a bistable multivibrator or similar circuit or device with two stable states. The flip-flop can have a number of inputs to switch from one state to the other state upon application of an input signal. The flip-flop usually has two outputs where only one is needed

to be used. A high level or "l" level output produces a low level output voltage, and a "0" level output produces a high level output voltage when the flip-flop is in a first reset state. The output voltage levels are reversed when the flip-flop is set to a second state.

A single sweep multivibrator is a monostable multivibrator or similar circuit or device which, when an "l" signal is applied to the input, where the input goes from "0" to "1" or from low to high level, produces a single output pulse with "l" level and with a selected duration. A delayed single stage multivibrator is a multivibrator which is such that an output pulse with "l" level and of selected duration is produced a predetermined time delay after an "l" signal is applied to the multivibrator's input.

An inverter is a device which produces a "0" output signal when an "1" signal is applied to its input, and which conversely produces an "1" output signal when a "0" signal is applied to its input.

It should be noted that when pressing the PFR signal

on the reset inputs of any of the conventional flip-flops, the flip-flops are reset. The reset of the flip-flops causes their output terminals, which previously showed a high level, to then show a low level. The reset also causes previously low-level inputs to go high.

In the following, the operation of the ultrasonic measuring device 20 according to the invention will be described with reference to fig. 4 and 5. The pulse repetition frequency circuit 30 is supplied with a pulse over the conductor 41 to the connection point 43 and then over the conductor 44 to the pulse device 48. The pulse device or the optical isolator 48 then supplies a trigger pulse to the trigger circuit 51.

The trigger circuit 51 supplies a positive potential to the control electrode 53 to cause ignition of the first controlled rectifier (SCRi 54 which supplies a large anode-cathode voltage to cause an ignition of the second SCR 63. Likewise, the controlled rectifiers 71 and 78 are ignited.

The four controlled rectifiers 54, 63, 71 and 78 are able to withstand the applied voltage which is evenly distributed by the resistors 84 and 86 - 88. However, when the first controlled rectifier 54 is made conductive, the remaining three rectifiers cannot withstand the applied voltage and is broken down as described above.

The successive ignition of the four series-connected controlled rectifiers 54, 63, 71 and 78 causes the capacitor 92 to discharge to impress a voltage pulse on the associated converter 21. The pulse impressed on the converter 21 is a negative pulse with a magnitude of 200 - 250 volts and a duration of approx. 60 nsec, and the pulse is supplied across the diode 96 and across the conductor 98 to earth. The supply of this pulse to the converter 21 causes the converter to generate pressure waves which affect the sections of the aligned sheathed cable 12.

An instant after the controlled rectifiers 54, 63, 71 and 78 are turned on, the voltage at the connection point 91 is low, essentially equal to zero, while the voltage at the connection point 93 is 200 - 250 volts. This causes the capacitor 92 to discharge. The capacitor 92 is chosen so that it is not discharged instantaneously, but instead is discharged over a certain period of time from the diode 102 through the connection point 98 and the diode 96. The capacitor is also discharged through the controlled rectifiers until there is essentially zero voltage across the capacitor.

From the source 109, current is supplied for recharging the capacitor 92 to the original state. Current in the controlled rectifiers is supplied from the rectifier 78 in the direction of the rectifier 54. Furthermore, the current drops from the peak value of several amperes to below this level, whereby the controlled rectifiers become non-conductive. It should be noted that this occurs after the trigger pulse has passed and before the next subsequent trigger pulse occurs.

In the following, the general operation of the receiver logic circuit 29 will be described. This circuit is designed to emit a positive pulse if the echo pulses received by the pulser/receiver unit 27 lie within a predetermined amplitude range. The receiver logic circuit 29 comprises logic circuits for performing outer band and inner band determinations for echo pulses. If the amplitude of the echo pulses received by the pulser/receiver unit 27 is above a predetermined amplitude, the circuit will register a negative pulse or outer band pulse.

A valid set of echo pulses received by the pulser/receiver unit 27 is spaced apart by a time interval 2t as shown in fig. 6, as the first echo pulse, denoted I, occurs approx. 40 - 120 sec. after the beginning of the cycle. The first echo pulse occurs when the pulse emitted by the associated ring converter crystal 21 affects the outward facing surface of the jacket 13. The second echo pulse, denoted II, occurs when the non-reflected part of the pulse emitted by the converter crystal hits the inward facing surface of the jacket 13.

The amplitude of the echo pulse I is a function of the acoustic impedance mismatch between the hot polyethylene sheath and the water in the cooling trough 14. The amplitude of the echo pulse II is a function of the acoustic impedance mismatch between the hot polyethylene sheath and the core 11. The first echo pulse also has a lower amplitude <p>g less duration than the second echo pulse.

During the "window width" (with a duration of approx. 90 - 120 ysec. see fig. 6h) a "hold-off" pulse (see fig. 6d) is generated by

- the receiver logic circuit 29 after receiving a first echo. A confirmation pulse (see Fig. 6e) starts immediately after the first "hold-off" pulse has ended. It is desirable that the first confirmation pulse occurs after the amplitude decrease of the first echo pulse. If the receiver logic circuit 29 indicates a positive output.158, it is known that a valid first pulse has occurred. Then, at the beginning of the second echo pulse, a second "hold-off" pulse is generated, with a second confirmation pulse beginning at the end of the second "hold-off" pulse. Unlike the first confirmation pulse, the second confirmation pulse should occur during the attenuation or decay of the second echo pulse. The part of the circuit that is connected to the other be-

force pulse, should therefore give a negative output 217 which indicates that the second echo pulse lies outside the predetermined limits of the threshold band.

It should be noted that if the receiver logic circuit 29 receives a negative output signal from an apparent first echo, the receiver logic does not wait for a second echo, and therefore does not generate a read pulse for the first cycle.

The confirmation pulse also avoids false capture of noise signals which could be due to air bubbles in the water medium in the cooling trough 16. If a noise signal precedes the first echo pulse, the circuit picks up the noise signal as a first echo and then tries to confirm the same first echo as an apparent second echo . If this were to occur, the receiver logic-. the circuit 29, which does not receive the expected positive-negative sequence, does not cause the read pulse to be generated.

The pulse repetition frequency circuit 30 also controls the function of the channels 22-22 and their interaction. To achieve this, the pulse repetition rate circuit 30 generates four pulse repetition rate signals, one for each of channels 22-22 and each with a period of about 1 msec. The four pulses are shifted approx. 250 ysec (see fig. 6j), thus causing a shift in the four channels' operation sequence. Thus, starting with the upper channel 22, and proceeding to the lower channel and then the left and right channels, all activity of significance occurs in the upper channel within the first 250 ysec, before the generation of the shifted pulse which is connected to the bottom channel.

The pulse repetition frequency circuit 30 adapts or prepares the circuit 29 for each cycle. A PRF pulse defined as a negative true pulse from the PRF circuit is applied to the high level or "l" inputs 143, 170, 222 and 240 of the respective flip-flops 141, 169, 219 and 242, to reset their outputs to low level or "zero" level. The application of the PRF pulse is considered the beginning of a sample cycle. The PRF circuit 30 also controls the circuit 29 so that it accepts echo pulses of the associated converter 21 only during the predetermined time interval designated as the window width (see Fig. 6b). The pulse from the PEF circuit 30 is supplied to the input 111 of the multivibrator 112.

The multivibrator 112 causes a window delay (see Fig. 6b) of a predetermined duration which is applied to the reset input 163 of the flip-flop 162 to cause its "0" output 164 to show a high level. The delay pulse is also applied to flip-flop 187's reset input 188 to cause its high level output 197 to go low and its low level output 189 to go high.

The trailing edge of the delay pulse applied across the conductor 113 causes the multivibrator 114 to generate a window width pulse (see Fig. 6).* The so-called "window width" of the window pulse is the time interval in which the channel 22 will receive valid pulse echoes from those emitted by the transducers 21-21 to the cable 12, thus avoiding random pulses. The end of the window pulse is considered the end of the sample cycle. At this time, the receiver logic circuit 29 determines whether to generate a pulse for continued control of the processing of measurements attributed to the reception of echo pulses.

The window pulse is emitted from the output 116 of the multivibrator 114 and supplied to the connection point 118 to enable the detectors 120 and 122.

The detection circuit 123 contains means for detecting one positive or one negative echo pulse which is supplied by the amplifier 106 to the input terminal 124. It should be noted from fig. 5 that the two threshold amplitude detectors. 120 and 122 are functional only when the window signal is present in connection point 118.

The threshold amplitude detector 120 is designed to supply a negative assertion signal to the NOR gate 131 input 129 in response to the supply of a valid positive echo pulse to the input terminal 124 of the threshold detection circuit 123. On the other hand, the threshold amplitude detector 122 is designed to also supply a negative assertion signal to the NOR gate's 131 input 133 in response to a valid negative echo pulse.

The threshold detection circuit 123 is designed to supply a signal to the NOR gate 131 only when the first echo pulse has at least one predetermined threshold amplitude. An echo pulse with a predetermined minimum amplitude is said to be out of band and causes one of the amplitude detectors 120 and 122 to generate a negative signal. If the amplitude is not at least equal to the predetermined threshold amplitude, the amplitude is said to lie within the band, and the negative signal does not appear.

When the polyethylene sheath 13 cools, the amplitude of an echo pulse from the sheath increases. The present system is designed to measure thickness and eccentricity as close to the extruder as possible. The measurement made will therefore apply to hot polyethylene material. The detectors 120 and 122 are selected and set so that they detect pulses having only a predetermined minimum amplitude corresponding to that expected at the outward facing surface of the polyethylene sheath 13.

The receiver logic circuit 29 then samples the first echo pulse to determine if the first echo pulse does not have more than a predetermined duration. Experiments have shown that the first pulse sent from the interface between the water and the outward-facing surface of the polyethylene sheath has a very short duration, for example of the order of 1/2 ysec. The second echo pulse from the interface between the inward-facing surface of the polyethylene and the core or screen layer, on the other hand, has a duration of the order of 1-2 ysec. This is due to the polyethylene's tendency to filter out high-frequency energy and the greater reflection at the other interface. Any signals that are to be valid must therefore meet criteria that have been established with regard to both duration and amplitude.

If the amplitude of the first echo pulse exceeds a predetermined value, the threshold detector 120 or 122 supplies a signal to the respective inputs 129 or 133 of the NOR gate 131. This in turn causes the NOR gate 131 to supply a positive or high level signal to the connection points 136 and 137 and to the inverter 138. The inverter 138 then supplies a negative signal across the conductor 139 to the flip-flop 141. Thus, the output 143 of the flip-flop 141 is set to a high voltage level to cause a high level to appear at the connection point 14 4 .

The flip-flop 141 can only be reset by applying a PRF signal at the beginning of the subsequent measurement cycle.

The high voltage level at the connection point 144 is supplied as a true input signal to the input 196 of the NAND gate 193. As shown in fig. 5, the delayed PRF pulse applied to the input 188 of the flip-flop 187 causes the output 189 to reset to a high level, thus causing a positive true signal to appear on the input 192 of the NAND gate 193. This causes the NAND -gate 193 generates a negative output pulse which signifies the beginning of the measurement cycle and causes the counter 32 to start counting.

The reason for the application of a delayed PRF pulse to the input 188 deserves a few words of explanation. If the output 143 of the flip-flop 141 should be at a high level and a PRF pulse is applied to the input 142 of the flip-flop 141 and to the input 188 of the flip-flop 187, the flip-flop 187 can react faster than the flip-flop 141, as that positive true signals would appear on both inputs 196 and 192 of NAND gate 193. This would cause a negative output pulse of short duration from NAND gate 193, but when flip-flop 141 is then reset, output 143 goes low which would cause a low level at the NAND gate 193's input 196 and thus interrupt the pulse output signal from the NAND gate 193 and thus the oscillator's pulse count.

To avoid this, a delayed PRF pulse is supplied to the input 188 of the flip-flop 187. Thus, the reset is delayed. the output 189 to a high level so that a signal with a high level is supplied to the input 192 of the NAND gate 193 only after zeroing the output 143 from the flip-flop 141 and the substantial setting of this by means of at least a partially valid ter - split pulse I.

The appearance of a high level at connection point 144 indicates at least a partially valid first echo. A check is initiated to verify or confirm this assumption by deciding that its duration is approx. 500 nsec., and that there is then a time interval or a gap where there is no signal at all. To this end, the high level at junction 144 causes a high level to also be applied to the input 146 of the first hold-off pulse multivibrator 147. The signal at junction 144 indicates the presence of a valid echo pulse insofar as the amplitude determined by a of the threshold detectors 120 or 122.

The application of a signal to the connection point 144 causes the multivibrator 147 to generate a time delay or "hold-off" pulse. (see Fig. 6d> to determine if the first pulse is of valid duration, eg 1/2 jsec., to qualify as a valid first echo pulse. The multivibrator 147

is designed to generate a "hold-off" pulse with a predetermined length which in this case is approx. 1/2 sec. After that 1/2 sec. the trailing edge of the hold-off pulse causes the confirmation pulse multivibrator 149 to generate a pulse (see Fig. 6e) with a duration of approx. 500 nsec.

During this time, the pulse causes a high level to appear on the NAND gate 154 input 153 to open the NAND gate. If the echo pulse has a duration that exceeds the "hold-off" pulse, a high level is applied across conductor 156 to NAND gate Trs^— 154 input 157 from connection point 136. This causes positive true input signals to be applied to both NAND gate inputs and causes a low level at its output 158.

If the echo pulse is still present, which would indicate that a valid echo pulse I has not been received due to too long a duration, the measurement of the signal is discarded. If, however, the echo pulse is not present, the signal is remembered.

The output of the flip-flop 162 is reset to a high level by the delayed PRF pulse which is applied to the input 163. If the NAND gate 154 is affected so that it has a low level at the output 158 which indicates an echo pulse which is longer than 500 nsec., the flip-flop 163 is set so that a low level appears at its output 164. The low level at the output 164 appears at the connection point 167 and at the so-called '^" input 168 of the flip-flop 169.

If the flip-flop's 169 "D" input has a low level when

an input signal on the clock or "C" input 171 occurs at the positive trailing edge of the acknowledge pulse generated by the multivibrator 149, a low level appears on the output 174. This corresponds to a no-gap situation, i.e. the first echo pulse is invalid. If the output 174 shows a low level, the NAND gate 178 is not operated and the output 174 remains at a low level. The next PRF pulse attempts to reset flip-flop 169. However, since output 164 is already low, the output remains low.

If, on the other hand, the echo pulse has disappeared after the "hold-off" pulse (which indicates a valid first echo pulse), no high signal occurs at the NAND gate 154's input 157. The NAND gate 154 is not operated, the NAND gate sets the flip-flop 162 and the high level of the output 164 appears on the "D" input 168 of the flip-flop 169. Since the "D" input is high when the clock signal appears on the "C" input 171, the output 174 is high, causing a high level on the NAND gate 178 input 177 to open the gate. Output 174 remains high until the next operation cycle when the PRF pulse resets flip-flop 169 and causes the output to be low.

If a high level occurs at the NAND gate 178 input 177, the first echo pulse has been confirmed with regard to time and duration. This allows the circuit 29 to wait for a second echo pulse.

The high level at junction 167 causes a high level at the NAND gate 227 input 228 so that the gate opens. The opening of the NAND gate 227 enables the gate so that if the second pulse received is acknowledged, the NAND gate facilitates the generation of a read pulse to cause storage of the counter 32 count.

When the second echo pulse is received by circuit 29 input terminal 124, one of threshold detectors 120 or 122 operates as described above to cause NAND gate 131 to generate a high level at its output 134. This causes a high level to be applied across conductor 179 and the connection point 181 to the already opened or functional NAND gate 178 input 182, so that the NAND gate is operated.

The actuation of the NAND gate 178 (due in part to the reception of the second echo pulse) causes a low level at the output 183 of the NAND gate to appear at the input 186 of the flip-flop 187 to set the flip-flop. This adapts or prepares the receiver logic circuit 29 to wait for the second echo pulse. This also indicates the end of the counting cycle (see fig. 6). The setting of the flip-flop 187 causes a low level at its output 189 and thus a negative signal to appear at the input 192 of the NAND gate 193 to disable the NAND gate and interrupt the generation of a low level at its output. This causes an interruption of the oscillator output count at counter 32 (see Fig. 6hX.

The remainder of the cycle is devoted to determining the validity of the second echo which would cause the transfer of the count to the buffer register (not shown). For the second echo pulse, a similar validation or confirmation process is used as that used to confirm the first echo pulse. As before: mentioned, the second echo pulse will have a duration of the order of 1/2 - 2 ysec. The part of the receiver logic circuit 29 which examines the second echo pulse is therefore designed with a built-in delay of 1 ussec. Only then is the pulse examined, and if the threshold amplitude is still present, the pulse is remembered.

The setting of the flip-flop 187 causes the low level of the output 197 to go high and is applied across the conductor 198 to the second "hold-off" pulse generator 201. The leading edge of the pulse from the flip-flop 187 causes the pulse generator 201 generates a pulse that has a duration of approx. 1.0 sec.

When the "hold-off" delay pulse is generated by the multivibrator 201, a signal is sent from the connection point 204 over the conductor 236 to the one input 237 of the threshold detector 238 to make this detector functional. The positive amplitude of the second pulse is checked by the threshold detector 238 to determine if the amplitude of the second echo exceeds a preset value that is much greater than the value of the original threshold. The one in fig. 5, the threshold detector 238 only checks positive signals, but it is also within the scope of the invention to incorporate another threshold detector in the circuit 123 to test the signal with regard to negative results.

In the event that the maximum threshold amplitude is exceeded, detector 238 applies a low level across conductor 241 to one input of flip-flop 242 to set the flip-flop and cause a high level to appear at its output 243. The high level at output 243 is applied across the conductor 244 to the input 236 of the multivibrator 233 to make the multivibrator functional.

The trailing edge of the pulse generated by the multivibrator 201 is applied to the second pulse confirmation multivibrator 208 and causes this multivibrator to generate a confirmation pulse (see Fig. 6g) of approx. 500 nsec. The second assert pulse multivibrator 208 causes a high level to be transferred from its output 209 across the conductor 211 and applied to the input 212 of the NAND gate 213.

From fig. 5 also shows that if either the positive or negative threshold is exceeded during the confirmation pulse, a signal is supplied from the connection point 137 over the conductor 179 to the connection point 181, and then over the conductor 216 and supplied as a high level to the NAND gate 213's input 214 .The ter discrimination test of the second echo pulse is necessary to determine if there is in fact any rash.

If high-level signals are applied to the NAND gate 213 input 214 from one of the threshold detectors 120 or 122, indicating that the second echo pulse is present, and to the input 212 from the multivibrator 208, a low level appears at the NAND gate output 217 for setting flip-flop 219. Setting flip-flop 219 causes a high level signal to be applied across conductor 224 to appear at input 226 of NAND gate 227 which has previously been enabled.

The NAND gate 227 is operated and then supplies a low level signal across conductor 231 to the single sweep multivibrator 233 which is designed to generate the read pulse and which was previously enabled by the amplitude confirmation.

The generation of a low-level pulse by the NAND gate 227 and its application to the multivibrator 233 (see Fig. 5) indicates that the first pulse had a duration of less than 1 ysec with a suitable gap to the second echo pulse, and that the second echo pulse had a duration of 1/2 - 2 sec.

After completion of the window pulse, a determination is made as to whether the read pulse should be generated. The end of the window pulse is indicated by a signal supplied by the monostable multivibrator 114 over the conductor 248 to the input 247 of the multivibrator 233. The multivibrator 233 which was previously made functional by a signal from the multivibrator 24 2 to the input 246, has now supplied an input signal from NAND -gate 227. The multivibrator 233 is now operated to command the counter 32 to register the pulse width from the NAND gate 193 upon receipt of the trailing edge of the window pulse on the input 247.

The oscillator 34 generates pulses which are counted by the counter 32 as this begins at a starting point of the first "hold-off" pulse and ends at the start of the second echo pulse, the difference between them being a measure of the time interval between the echo pulses. The counter 32 comprises three decade counters, hundreds, tens and ones, a 4-bit memory being linked to each of the decade counters. If correct pulses are generated with respect to echo pulses I and II

in a positive-negative sequence, a read pulse is generated at the end of the window pulse's "window width". This causes the count stored in the counter's 32 register to be transferred from the counter's counting part to the counter's memory. Then the count from the memory is applied to the digital-to-analog converter 36 to influence the digital-to-analog converter to provide a continuous output voltage indicative of the cladding thickness.

It should be noted that the previous count stored in the buffer register is not updated until the next subsequent, valid count pulse is generated to transfer the corresponding count to the buffer register. The circuit 25 is designed so that a previous count is retained in the buffer register until the next valid count is received, and is not discarded if an invalid count is performed.

The digital-to-analogue converter also contains a lamp which is operated for approx. 800 ysec. between successive read pulses.

Four of the lamps are connected to the device 20. If acceptable measurements are taken, the lamps will be in the on state approximately 80 to 90% of the time. If, on the other hand, the signal is lost, there will be no light at all, which may be an indication that a crystal 21 is out of position or that a channel is defective. The lamp for each of the channels 22-22 is used as a warning lamp for the operator if the system fails and also during installation of the equipment. The lamp is continuously on if a continuous stream of read pulses is generated. An 80% duty cycle signal would drive the lamp under such a condition.

When a cable swings through the crystal's field of view, the lamp will

tend to flicker as no excitation will be present when the crystal cannot "see" the cable.

The device 20 has certain features, which can be traced to the thickness of the jacket 13 being tested. Coats with a thickness of at least 0.5 mm usually do not cause any problems. If, however, the sheath thickness is less than this, problems can arise. The receiver logic circuit 29 is designed with respect to the first "hold-off" pulse generation and confirmation part, to be adjusted to 0.5 mm per unsec.

For example, the hold-off pulse multivibrator 147 must generate a hold-off pulse at least equal to the swing damping time of the associated crystal 21. It is possible to select a crystal that has a damping time as low as 0.5 ysec.

Furthermore, the confirmation pulse generator 149 must generate a pulse with a duration that is at least equal to one oscillation cycle for the associated crystal 21. The crystal 21 can oscillate with a frequency of 5 MHz so that the pulse width can be as low as 200 nsec.

Once these two durations are determined, successful operation of the device 20 is dependent on the second echo (echo II) not being received before the completion of the first hold-off and confirmation process for the first echo. It is possible that this requirement cannot be met if the device 20 is used to measure sheaths with a thickness of an order of magnitude smaller than 0.5 mm.

With some of the more commonly used insulating plastic materials for thinner casing walls, the time between echo pulses is of course longer due to the fact that the speed of the ultrasound wave in this particular material can be lower. In hot polyethylene, the expected speed is 0.51 mm per <y>sec. Working with a polyethylene sheath that is thinner than e.g. 0.5 mm, can thus involve problems with regard to echo pulse overlap. However, for a given thickness in hot polyvinyl chloride, the time duration between echoes I and II can be longer, so that any overlap is avoided even if the sheath is thinner.

One will of course realize that it is within the scope of the present invention to measure the thickness between a first surface and a second surface which is separated from and lies opposite the first surface with several layers of different materials in between. Instead <*>to measure the thickness of a cable sheath, the method and device according to the invention can also be used to measure the thickness of successive sections of a massive core, or a full core such as e.g. by pipelines.

It should be noted that in some cable constructions where

e.g. an inner mantle is engaged with a core envelope or other material, the second echo pulse for each associated echo pulse pair may be extended. However, it will be remembered that the generation of the off signal by means of the NAND gate 193 occurs after receiving the initial part of the second echo pulse for the purpose of performing thickness measurement. Of course, the peak size and duration characteristics for the second echo pulse will also be satisfied. If, on the other hand, the core envelope is separated from the inward-facing surface of the mantle, the air acts as an open barrier.

ter through which, in the main, no transfer takes place. The testing and application of the second echo pulse in thickness measurement

the gene is not degraded here either.

Claims (5)

1. Procedure for ultrasonic measurement of the thickness of an elongated element where relative motion exists between successive sections of the elongated element and a source of electric ultrasonic pulses, and first and second echo measurement pulses from different, opposite surfaces of the element are distinguished from noise and unused echo pulses by a qualifying procedure comprising the steps of sending an ultrasonic pulse into the element, receiving the first echo pulse with a minimum amplitude in response to the ultrasonic pulse from one surface of the element, receiving the corresponding sending a second echo pulse with a minimum amplitude from an opposite surface of the element within a given time period, and developing signals after receiving the first and second echo pulses with a predetermined minimum amplitude, the time duration between the signals being related to the elapsed time between the reception of the first echo pulse from one surface of the element and receiving the associated second echo pulse from the opposite surface of the element, characterized in that the first echo pulse (i in Fig. 6a) is confirmed by testing with respect to a first predetermined timing characteristic comprising duration in addition to minimum amplitude, and the second echo pulse (II) is confirmed by testing with regard to a second predetermined timing characteristic comprising duration in addition to an amplitude that is greater than the minimum amplitude, the signals developed upon receipt of the first and second pulses being respectively distinguishable on and off signals, the on and off signals being developed only in response to the reception of first and second echo pulses having at least the minimum amplitude, and the off pulse being developed only if a predetermined time interval has passed after the beginning of the first echo pulse. 2. Method according to claim 1, characterized in that the generation of a control pulse (e in Fig. 6) is partially conditioned or prepared after confirmation of the first echo pulse, and that the conditioning of the means to enable the generation of the control pulse is completed in response to the confirmation of the second echo pulse, and the measurement of the elapsed time is processed in response to the generation of the control pulse. 3. Method according to claim 2, characterized in that the test characteristic for completing the confirmation (e in Fig. 6) of a qualifying first echo pulse is a duration of the qualifying test characteristic which is less than a preset time. 4. Method according to claim 2 or 3, characterized in that the test characteristic for completion of the confirmation (g in Fig. 6) of the associated second echo pulse includes a peak amplitude and a duration that is greater than preset values. 5. Device for ultrasonic measurement of successive sections of an elongated element according to the method specified in claim 1, comprising a device for sending an ultrasonic pulse to the element and then receiving a first echo pulse which is reflected from one surface of the element and then receiving an associated second echo pulse from the opposite surface of the element, and a device (120, 122 in Fig. 5) for qualifying the first echo pulse and then the associated second echo pulse based on amplitude , - characterized in that it comprises a device (29 in Fig. 3; 238, 187, 193, 147) for, in response to the receipt of a qualifying first echo pulse, to initiate development of an on signal and examination of a confirmatory test characteristic for the first echo pulse, a device (178, 227 in Fig. 5). for, in response to the confirmation of the first echo pulse and the receipt of a qualifying associated second echo pulse, to initiate examination of confirmatory test characteristics for the qualifying associated second echo pulse, a device (187, 201, 238, 208) for cause the development of an off-signal where the time duration between the on-signal and the off-signal is related to the elapsed time between the reception of the first echo pulse from one surface and the reception of the corresponding second echo pulse from the opposite surface, a device (32, 34) which measures the duration of time between the on and off signal, a means (169) for, partly in response to the confirmation of the first echo pulse, providing means to enable the generation of a control pulse, ,. a device (227, 233) for completing the preparation of the enabling means in response to the confirmation of the associated second echo pulse to generate the control pulse, and a device for processing the measured duration in response to the generation of the control pulse.<6>* Device according to claim 5, comprising a comparison device (123 in Fig. 5) for testing the amplitude characteristics of the first and second echo pulses, characterized in that it comprises a device (112, 114, 20, 122 in Fig. 5) for testing the duration of a first pulse with a required amplitude characteristic, a device (122) for testing the duration of an associated second pulse, the testing of the duration of the second pulse starting the comparator (123 in Fig. 5) to test the peak amplitude for the second pulse, a device (162, 169) for developing the on-signal after receiving the first echo pulse which has the required amplitude, and the off-signal after confirmation of the first echo pulse and during the reception of v a second pulse, a first logic device (238, 147) responsive to receiving the first echo pulse having the required amplitude, to start the development device and the first pulse duration device, a second logic device (187) responsive to the valid first echo pulse and the associated second echo pulse having the required duration, to enable an influencing device (193)., a third logic device (178) which responds to the first echo pulse to enable the second logic device (187). for developing the off-signal and for influencing the second pulse confirmation device (201, 208), the confirmation of the second pulse with regard to peak amplitude further making the influencing device functional, and a device that starts a test cycle for operation of the influencing device when this is made functional at the end of the test cycle, and for preparing the circuit for a further operating cycle.