DE7638775U1 - Eddy current tester - Google Patents

Description

PATENTANWÄLTE I>'' · "· * ·

ι VUESTMOFF

WUESTHOFF-v. PECHMANN-BEHRENS-GOETZ d, pl, in _O .cerhard pu,

DIPL.-CHEM. DR. E. BARBER OF PECHMANN PROFESSIONAL RBiRBSENTATIVES BEFORE THB BUROPEAN PATENT OPPICB BR.-INC. DIETER BEHRENS

MANDATAIRES AGRE £ s PREC!. 'OFFICE EUROPEEN DBS BREVBTf DIPL.-ING .; DIPL.-WIRTSCH.-1NG. RUPERT GOBI

D-8000 MUNICH 90

. G 76 38 775.2 SCHWEIGERSTR / SSE2

Magnetic Analysis
1G-48 830 phone: (o8 j) 66 io j ι

telegraum: protectpatent TEL7.X: 5x4070

December 12, 1980

Description:

MAGNETIC ANALYSIS CORPORATION 535 South 4th Avenue, Mount Vernon, New York 10550

Eddy current tester

The invention relates to an eddy current testing device according to the preamble of claim 1.

Eddy current testing is primarily based on changes in conductivity to identify cracks and defects or, more generally, any irregularities that prevent the flow of Change eddy currents. For the sake of simplicity, the term "cracks" will hereinafter be used to include similar errors and Includes changes. Eddy currents are induced in an object to be tested. Changes in eddy currents will be detected and displayed. When testing magnetic material, irregularities call the permeability of the material A background noise emerges that covers up changes in conductivity caused by cracks and the like. Such Changes in permeability can be based on local stresses and hardening, among other things. Relatively large cracks can never be detected small cracks remain undetected. But it is precisely such small cracks that can be very important, especially if, for example, the condition of a pipe is to be determined before it breaks.

/ 2

-2- "*" 48 830

It has already been suggested to saturate the magnetic Materials to use a magnetic field generated with direct current, thereby reducing the influence of permeability changes turn off. In a known eddy current test device according to GB-PS 956 053 are continuously with DC powered coils are used. Such coils will overheat once they are small enough to fit in a pipe to be pushed in. This overheating occurs in particular occurs when a strong current is used with materials that require a high field strength to be saturated must become.

^ The invention is based on the object of a generic Eddy current testing device to be developed in such a way that a test of pipes and the like for small cracks and irregularities is also possible from the inside.

This problem is solved with the features of the main claim.

In the eddy current according to the invention, it is possible to use a To train the test head or a probe so small that he or she can be inserted into pipes, which are tested in situ can be. For example, probes with a Diameter between 1.5 and 5.7 cm, field strengths of 1000 Oersteds and more are achieved. With eddy current test devices according to the invention This means that pipes can be tested without having to remove the top plates of the pipes. at Plagued by cracks, only the pipe with cracks needs to be replaced. There are regular tests of pipes possible while the pipes are in operation.

-3- 48,830

In order to determine the current that leads to sufficient saturation of the pipe or test object, preliminary tests can be carried out with be performed on a test specimen that has an artificially generated crack. The current can also be used under test conditions can be increased until the background noise is reduced to a minimum. It is advantageous to have a continuous To have an indication that the test item has been brought into its saturation range.

Advantageous embodiments of the invention are set out in the subclaims marked.

The invention is explained below with reference to schematic drawings, for example and with further details.

They represent:

Pig. 1 a probe for internal testing of a pipe,

Pig. 2

and FIG. 2a shows an embodiment of a probe as it is in a device according to the invention is used, Pig. 2 a cross-section along line 2-2 of Pig. 2a represents

Pig. 3 typical normal magnetization and permeability curves of a magnetic material,

Pig. 4 composite DC voltage pulse waveforms,

Pig. 5 is a block diagram of an eddy current testing device, j

Pig. 6 one in the device according to Pig. 5 usable pulse generator,

Pig. 7 waveforms of the generator according to FIG. 6,

Pig. 8 one in the device according to Pig. 5 usable impulse drivers, ||

"": !! . ''. ''. · '·· [ "''" / follow original pages ^8': .. · '· .. ^{:: ..:::} ki ^ 32 unchanged

6764 l ^l

- 8 - 48 830

Fig. 9 waveforms of the pulse driver according to Fig. 8,

Fig. 10 is a CRT (cathode ray oscilloscope) representation of the output of the tuned amplifier according to Fig. 5,

Figure 11 is a CRT representation of the differentiated output of a single take-up reel,

12 shows a modified composite DC voltage pulse,

Fig. 13

and 14 other pulse waveforms for testing magnetic materials

15 is a block diagram of a modified crack detector with a bridge circuit;

16 shows an apparatus of the comparator type which is used with composite Impulses works,

17 shows a modified embodiment of a device of the comparator type having a bridge circuit and FIG

Fig. 18 shows a comparator type apparatus having an electrically generated reference signal.

1 shows a pipe 10 to be tested by means of an inner probe 11, for example a boiler or heat exchanger pipe, which is mounted in support plates 12. Typically, such pipes are made of a magnetic material, such as low carbon steel or magnetic stainless steel, and a large number of pipes are mounted in the walls. The probe 12 is intended to be inserted into a pipe in situ so that the condition of the pipe can be checked without removing it.

Figs. 2 and 2a show a test probe. A primary coil 13 is around a layered core 14 is wound. Preferably a core material chosen that its saturation at a higher flux density than is required for the pipes to be tested, for example Vanadium Permendur. One with an end cap

/ 9

- 9 - 48 830

Screw bolts 15, for example made of pig iron, with a screwed on End cap 16, holds the core together. A drawn-in recess allows a cable to be easily attached; 17th for moving the probe 11 through the pipe under test 10. The primary coil 13 is surrounded by a pair of zero detector coils 18, 18 ′ which are wound around a bobbin 19. The probe 11 can be enclosed by encapsulation material 21 according to FIG. 1 in order to avoid damage during use.

A region of the pipe 10 to be tested is indicated by 10 'and a flow path through the core and the pipe wall is indicated by 22. Narrow air gaps are normally formed between the probe and the inside diameter of the pipe because it is necessary to have the Allow debris, debris, etc. to accumulate in the pipe. Overall, the probe is designed to be so effective in the pipe wall a high flux density is generated as much as possible.

In a special embodiment of the probe, there is the primary coil from about 40 turns and is worked with peak currents - of 100 amperes and more, which leads to the generation of the magnetic field over 4000 ampere turns are available. The number of ampere-turns required depends on both how light it is or it is difficult to saturate the pipe material as well as the probe dimensions, pipe wall thickness, etc. Moisture strengths can occur of up to 18-21 kilogauss may be required. For current applications, there is a design criterion for field strength the probe at 1000 Oersted, but this criterion can vary considerably.

In the case of high currents, it can be advantageous to provide a device for cooling the probe, for example for air cooling.

Fig. 3 shows a typical normal magnetization curve 25 of magnetic material. Be ^ giringen magnetic field strengths H is the Flux density B is small and increases. The flux density then increases significantly with increasing magnetic field up to the break point 26 or a significant curvature of the curve 25 is reached, whereupon the curve

/ 10

- ίο -

flattens out in the saturation area. The curve shown is plotted semi-logarithmically in a conventional manner, so that the large magnetization area is visible. Plotted linearly, the kink point is much sharper.

The dashed curve 27 indicates the permeability (B / H). The permeability depends on the flux density and is in general at low flux densities small, at medium flux densities maximum and then increases again in the saturation range of the material to a small value.

When a magnetic field generated by direct current

an alternating magnetic field is superimposed, the permeability effective for the alternating magnetic field can be referred to as "differential permeability" and defined as / 1 B / Δ H. In eddy current testing, it is advantageous to avoid changes in permeability, so it is desirable that the differential permeability approximate unity. The differential permeability depends on the size of the alternating current force or the alternating field, the size of the direct current force and the magnetic properties of the material as well as its history. In general, the larger the DC magnetic field, the smaller

is the differential permeability. Also applies if that: Equal rommagnetfeld close to the inflection point of the B-H curve is that the differential permeability is smaller, the smaller is the AC amplitude. '

It is currently preferred to have a sufficiently large DC magnetic field to work to bring the material into its saturation range, where its permeability compared to its maximum permeability is small. The latter is represented by the apex of curve 27 in FIG. 3.

The shape of the curves shown in Fig. 3 may vary depending on the magnetic material and the strength of the saturation required Magnetic field be noticeably different. In some

1 I 4 * »

3D - 11 - j. ; , 43 pity ":

In some cases it can be difficult to generate a flux density sufficient to reduce the differential permeability to 1 without the probe being unduly heated and destroyed. However, by creating a flux density sufficient to reduce the permeability to a relatively small value, a significant increase in the sensitivity of the tester to small cracks can be obtained.

Figure 4a shows a series of composite pulses for application the driver or primary coil of a Prüfkoplsonde, like them is currently preferred. Each composite pulse 30 contains a broad DC voltage pulse 31, which is a series short DC voltage test pulses 32 is superimposed, the period of which is short compared to the duration of the composite pulse. After a noticeable delay, the composite pulse occurs again, as shown by broken line 33.

The wide pulse 31 is a DC or unipolar pulse between a reference potential denoted by zero, generally the earth potential, and a higher potential V. Insofar if it is a direct voltage pulse, it generates a direct current in the coil and correspondingly in the object next to the coil a magnetic flux of uniform direction. Because the coil has an inductive resistance, the current increases during the Impulse as indicated by the dashed line 34. At the end of each pulse the current falls as if through the dashed line Line 35 indicated. The inductive resistance of the coil is determined in part by the magnetic material being tested and accordingly, the shape of the current curves depends in part on the material tested. Advantageously, the peak value and the length of the pulse selected so that the material reaches its saturation range before the end of the pulse. If the If the permeability changes, the shape of the current waveform will also change and when the permeability reaches 1, a Kinks occur after another slope is present. The curves 34 and 35 thus only represent a general trend.

/ 12

■ ··

The series of test pulses 32 enables cracks to be detected in an object under saturation conditions. Their frequency (PRF) can be based on factors known from eddy current testing to get voted. For example, 2.5-80 kHz can be used. As far as the impulses are above the level of the wide pulse, they do not reduce the DC magnetization caused by the wide pulse. Because the test pulses also have a DC voltage component given by their duty cycle, they contribute something to the DC magnetization at. For example, pulse duty factors of 25, 50 or 75% can be used, which correspond to percentages of the DC voltage components results. Other pulse duty factors can also be used.

After a period of time, the composite pulse repeats, as indicated by 30 ', so that the composite pulses are fed to the primary coil intermittently. The overall duty cycle can be selected to be sufficiently small that an impermissible one Heating up and possible destruction of the probe can be avoided. It was successful with a duty cycle worked from 12 to 1/2%, which results in an average heating current of about 12 to 1/2% of the peak or peak current.

In Fig. 4a only four test pulses are shown, in practice a much larger number can be used. For example In one test the duration of the composite pulse was 12.6 ms and the test pulses were a group of 10 kHz Pulses that were present during the second half of the composite pulse. That way were instead of the four test pulses shown, about 60 test pulses are available. However, this is difficult to represent.

Figure 4b shows a similar composite pulse in which the peak values of the test pulses 36 equal the initial value 37 of the broad impulse are. The minimum values 38 of the test pulses are greater than the minimum value 39 of the composite Impulse. Here the DC component of the composite pulse is somewhat reduced during the latter part,

where the extent of the reduction in the duty cycle of the test pulses, the duration of the test pulse group and the minimum value 38 of the test pulses depends. In many cases this can decrease but be insignificant.

Fig. 4c shows another example in which the test pulses 41 occur at the end of the broad pulse 42. The combination can be viewed as a composite pulse 43 which recurs intermittently, as indicated by 43 *. Although the DC component of the test pulses is smaller than the of the wide impulse, satisfactory results are obtained in many applications. The inductive resistance of the Coil to which the pulses are fed counteracts the drop in current after the end of the pulse component 42 and each Pulse 4Ί partially restores the current. This can be done for Allowing a test in which the magnetic material is still in its saturation range, or at least at least, is sufficient is magnetized enough that its permeability is reduced to a sufficiently small value.

In describing the waveforms, the minimum and maximum values, etc. have been explained so far. The actually the coil applied waveforms can have the polarity shown or the polarity reversed, with the minimum, maximum and similar values are related to zero reference potential (usually Erdo).

Fig. 5 is a block diagram of an eddy current testing device according to the invention. With 51 an oscillator and square wave generator is referred to. The oscillator is advantageous a sine wave generator tunable with the aid of an oscillating circuit 52, the sine wave being converted to a square wave which is on the line 53. The square wave is then divided in divider 54 so that the desired PRF of the test pulses arises. A selected output 54 is fed to a further divider 55 which produces the initial wide pulses of the desired

/ 14

Duration generated. The output of divider 55 is to an initial pulse generator 56 connected, which generates wide DC voltage pulses of desired duration and desired duty cycle. The outputs of 54 and 56 are connected to a test pulse generator 57, the groups of test pulses to the desired Dividing the initial broad impulse is generated. The outputs of 56 and 57 are fed to a pulse driver 58 which the generated composite pulses shown in Fig. 4a. The composite pulses are fed to a coil driver 59 and from there the primary driver coil or primary coil 13 of the probe fed.

The DC component of the composite DC pulse generates a magnetic flux in the wall of the pipe to be tested, which is preferably sufficient to keep the wall in its Bring saturation area. The alternating voltage component creates eddy currents in the wall. Changes in eddy currents due to cracks or other defects in the wall will be zero detector coils 18 and 18 'added. In the example shown, the primary and zero detector coils are for simplicity shown outside the tube 10, but actually located inside the tube as shown in FIG.

The outputs of the zero detector coils 18 and 18 ′ are fed to an input amplifier 62 via a transformer 61. The input amplifier 62 is, as shown by the voting circle 63, tuned. Usually the amplifier is on the PRF of the test pulses Voted. In some applications, however, it is advantageous to set it to a harmonic of the PRF or one to the pulse width related frequency as described in U.S. Patent No. 3,786,347.

The output of amplifier 6 2 is connected to a pair of quadrature detectors 64 and 64 '. In a gate or pulse generator 65 with the help of the test pulse generator. ^r > 7 gate

/ 15

- 15 -

'., 48

or pulse pulses generated for the quadrature detectors 64 and 64 ', which correspond to the test pulse series. These key pulses are generated with the help of a signal from the start or trigger pulse generator 56 limited to desired intervals.

As can be seen from Fig. 4a, the test pulse groups are intermittent. As a result, a transition state occurs at the beginning of a given group. In the tuned resonant circuit 63 of the amplifier transient conditions also occur. To switch off transitional responses and display signals only then when a steady state is reached, the outputs of the quadrature detectors 64 and 64 'become a pair of second Detectors 66 and 66 'supplied. These second detectors 66 and 66 ' are gated by signals from a second strobe pulse generator 67 at times that are consecutive to the beginning Test pulse groups are delayed.

In the preferred embodiment described below the gating occurs at the end of the composite signal according to FIG. 4a, so that for the quadrature signals to achieve stationary values a maximum time is available. This also enables the PRF of the test pulses and the duration of the composite Pulse to change without an adjustment of the gating of the second detectors is necessary. The second Detectors 66 and 66 'can be read and hold detectors so that their outputs remain constant until they pass through a Change the following test pulse group. When the quadrature detectors If necessary, the second detectors can hold their output signals at the end of the respective test pulse group be gated or gated after the end of a composite pulse.

The resulting quadrature signals from detectors 66 and 66 ' are fed to amplifiers and filters 68 and 68 'and then to a display device 69. This can be done in a conventional manner. The display device 69 may be a cathode ray tube which displays both the amplitude and phase of the crack signals.

- 16 - ':: 48L 830 *. ^:

lit Si

Furthermore, alarm circuits etc. can be provided if necessary.

Fig. 6 shows the part contained in the dashed box 60 of Fig. 5 ent. The letters in Fig. 6 denote in Fig. 7 posed waves.

In Fig. 6, the frequency of the oscillator 51 is higher than the PRF of the test pulses and is chosen to allow conventional division to generate both the test pulse frequency and the duration of the wide initial pulses. The frequency is chosen here to be 320 kHz. The output of the oscillator is fed to a binary {counter-divider 54 having a plurality of outputs 71 for the desired range of PrüfImpulsfrequenzen. One of the outputs of the counter divider 54 is selected by a multiplexer 72 under the control of inputs a, b, c. An 8: 1 multiplexer is shown here which allows any of eight inputs to be placed on output line 73. Test pulse frequencies of 5 to 80 kHz are currently preferred, but lower and higher frequencies can be used if necessary.

To simplify the representation of the waveforms, a Frequency of 1250 Hertz was chosen. The corresponding frequency on line 73 is 2500 Hertz and is shown at C. These The frequency or waveform is inverted to form wave D. These waves are fed to associated flip-flops FF1 and FF2, ι whose function is to divide it by 2 and create a 90 ° - H to produce phase shift between their outputs.

The flip-fiops can be of a commercially available design. As shown here, the positive deviation of the wave with respect to the clock pulse CL transfers the state at the D input to the Q output. Q is the inverse of Q and R is a reset terminal. Initially are these Flxp-Flops and the other flip-flops in the Drawing reset so that Q is low (0) and Q is high (1). The counters are also reset. This is done by a circuit 80 reached, in which the capacitor is charged when the switch is closed and as soon as a predetermined voltage is reached

is, the output R of the inverter goes to a low level for resetting.

FF2 divides input C by 2 to create wave F. The connection between FF2 and FF1 delivers together with the Input D has an output E, which has the same frequency as F, but is phase delayed by 90 ° compared to F. One exit G of the divider 54 is fed to a further divider 55, some outputs of which are fed to a multiplexer 74, so that a choice of the duration of the initial pulse is possible. An output H with 78 Hertz is selected here.

So that there is room for subsequent waveforms, the wave H is moved by half a cycle to the left into the position H ¹ , as shown in dash-dotted lines.

The output H ¹ is fed directly to a NOR element 75. It is also fed to a decade counter divider 76 which generates a pulse J for every tenth pulse H. In counting, divider 76 holds its output the full cycle of high so that the width of the pulse in J is equal to one full cycle of high. The pulse J is inverted and forms K, which is fed to the NOR gate. The NOR gate operates in a conventional manner and gives a low (0) output when one input is always high (1) and a high (1) output when both inputs are low (0). The inputs H ¹ and K of the NOR gate 75 result in an output L which is high when H 'and K are low.

The interval between successive pulses in J and K is due to the division by 10 in the decade counter 76 Nine times the pulse length. This is difficult to illustrate in FIG. 7. The interval, however, is designated by 33 in FIG. 4a.

Wave L is fed to FF3 and FF4 to produce outputs M and N which are on the positive ramps of their clock inputs supplied waves F and E go to high levels. The outputs M and N remain until the positive increases of F and E

- 18 - . \ '.,' 48. .83D I *., ·

high following the point in time when L goes low. The outputs are used to correct their timing Generate groups of quadrature gates and test pulses.

The inverted outputs M and N are fed to a three-pole three-position function switch 77. In the illustrated Positions, R is fed to the NOR element 78 together with the shaft E, creating an output V that is inverted to E but only appears within the lower deviation of H, which corresponds to the upper deviation of N, as shown. Similarly, the wave M becomes the NOR gate 79 along with the wave F supplied, creating an output U that inverts to F. but occurs only within the short duration of M which corresponds to the positive duration of M as shown. The waves U and V are phase-shifted by 90 ° in terms of their frequency, so that quadrature gate signals for the quadrature detectors 64 and 64 'of FIG. 5 are provided. The signals result after differentiation, short gate or key pulses on their edges. Respective pairs of the pulses with a 90 ° phase shift are then used in a known manner for gating the detectors. In a special embodiment, differentiated impulses were used, which were sent to the positive-going edges of U and V occur.

The wave M is fed to a NOR gate 81 and generated together with the NOR gate 82 groups of test pulses that are in the correct temporal relationship with the broad initial pulse, such as shown in Fig. 4a. With switches 83 and 84 running at the same time, three different pulse duty factors can be selected for the test pulses will. In the positions shown, pulse duty factors of 50% are generated. One input of NOR gate 82 is grounded. on in this way the output is P and is fed to NOR gate 81 together with M. Correspondingly, the exit at 85 is the one Part of F that is present during the lower part of M, which corresponds to the high region of M as shown. the emerging groups are shown at T-1. As in the further

/ 19th

- 19 - 48 'Ü'3'cJ · I

will be explained, the pulse driver uses the lower (O) regions of the test pulse waveform to generate appropriate ones Driving impulses. Similarly, the wave is inverted at 85 to produce the output test pulse group at 86.

Both waves E and F are fed to the NOR element 82 for pulses with a pulse duty factor of 75%. The output of the NOR gate 82 is low when E and F are low. This is inverted by the NOR gate 81 and to the lower range of FI limited. The test pulse group arising in line 85 is shown at T2.

For pulses with a duty cycle of 25%, the waves E and F are also fed to the NOR element 82, which becomes the output inverted by the inverter 87 before it is fed to the NOR gate 81. The output is accordingly in the line 85 here inverted to the output for pulses with a duty cycle of 75%, as shown at T-3.

The wide initial pulse is generated by NOR gate 88, the inputs of which are J, N and M. The output P is low when either of the inputs is high. Similarly, the initial pulse P goes down when J goes up and stays low until N goes down.

The end of the positive displacement of N occurs slightly after the end of M. Because H is used to generate the test pulse group the end of the initial pulse P is a little after the last test pulse. Furthermore there are the gate or test pulses U and V in terms of time so that by selecting the appropriate polarities of the differentiated impulses generated from them, the last Gating of the quadrature detectors takes place before the end of the initial pulse. Any additional pulses generated at the end of P. therefore do not affect the proper functioning of the circuit.

/ 20

The function switch 77 in its middle position B grounds an input of the NOR gates 78, 79 and 81. As a result, a continuous series of test pulses is generated at the output 86 and are the. T _as t: .- Pulse outputs U and V continuously. This enables the device to be used for testing non-magnetic materials where composite pulses are not necessary.

In the calibration position of the function switch 77 there is an input of NOR gates 78 and 79 grounded, so that continuous series of pulse pulses are generated. One input of the NOR gate 81 is supplied by a counter divider 89 so that different lengths of test pulse groups can be obtained for calibration.

Fig. 3 shows a circuit for the pulse driver 58 and the second. Touch pulse generator. 67 within box 91 of FIG. 5. FIG. 9 shows, for explanatory purposes, waveforms associated with the corresponding numbered digits in Fig. 8 occur during normal operation.

The initial pulse waveform P shown in Fig. 7 is supplied to the input line 92 and then through a diode D1 and a resistor to the base of transistor Q1. Q1 is connected as an emitter follower and its output voltage is provided by resistors R1, R2 divided and fed to the base of transistor Q2. The supply voltage V2 is through a Zener diode 93 at a desired Voltage kept below V.

At time t ^ the initial pulse is high, Q1 is on (permeable) and Q2 is on. If the initial pulse at time t., Low Value goes, the transistor Q1 is switched off as well as Q2. The voltage V3 is normally taken from the supply voltage line with the voltage V supplied via Q9, the function of which will be explained later. The resistor R3 in the collector circuit from Q2 produces a preselected amplitude of the initial pulse as shown at 4. This is via the diode D2

/ 21

- 21 - · "'4b'

fed to the bases of transistors Q4 and Q5. Through the resistances A voltage division occurs between R3 and R5, which is selected here for representation as pV3.

The test pulse group from line 86 of FIG. 6 becomes the basis fed from Q6 which works as an emitter follower. Part of the emitter output, determined by the voltage dividers R6, R7

is fed to the base of Q3. Q6 is initially on and Q3 is I also on. At time t _2, Q6 and Q3 switch off and then go on and off according to the test pulses.

The collector output of Q3 generates a test pulse group as shown at 3 in this way. This test pulse group is fed to transistors Q4 and Q5 via diode D3, the amplitude of R4 and R5 being voltage-divided. R4 is smaller than R3 and the voltage division for the test pulses is assumed here to be 4v3 for the sake of illustration. While Q2 is off between t ₁ and to, the higher voltage at 94 biases D3. If, however, Q3 runs out (becomes impermeable) between t2 and t3, the higher voltage of shaft 3 passes through D3 to point 94 and preloads D2. Thereafter, the conditions change alternately for successive test pulses.

The waveform at the bases of Q4 and Q5 is shown at 5 in FIG. This waveform is the same as that of the composite Pulse according to FIG. 4a, the amplitude of the superimposed test pulses 3 2 25% of the amplitude of the broad pulse 31 is. The emitter outputs of Q4 and Q5 are fed to the coil driver 59 of FIG. 5 via a line 95.

In operation it is desirable to be able to regulate the current in the primary coil of FIG. Resistor R8 is in series for this purpose connected to a transistor in the coil driver 59, to which the composite pulses on the line 95 are applied, whereby a voltage proportional to the current in the stage is generated.

/ 22

-Δ

- 22 - · 4δ 83ϋ:. · * ·: * Γ ":

III | t (it- * ·. * * · ■ - »

A desired fraction of this voltage is taken from potentiometer 96 and applied to the base of Q7. Obviously this voltage is below the voltage in the voltage supply line 97. The diode D4 protects against a possible polarity reversal. The output voltage of Q7 is divided by R9, RIO and fed to the base of Q8. The coil output from Q8 is fed to the base of Q9. In normal operation, Q7, Q8 and Q9 are all conductive and will the conductivity of Q9 is controlled by the voltage taken from potentiometer 96. This holds the voltage V3 at a desired value. A large capacitor 98 is parallel to the base of Q9 so that the conductivity of the pulse increases Pulse is constant, but changes with the average current supplied to the driver stage.

For safety, the maximum current is passed through the coil driver stage limited by a circuit 99 to avoid damage. The circuit here is more optical than commercially available Insulator 101 executed. Roughly speaking, controls a light-emitting one Diode a photosensitive transistor. The diode responds to the voltage drop across resistor R8. if this voltage is too high, the transistor becomes conductive to diminish the input of Q9, causing V3 and accordingly the amplitude of the composite pulse in line 95 is reduced.

The gate pulses for the second detectors 66 and 66 'of FIG are also generated in the circuit according to FIG. In addition becomes the emitter output of Q1 through R11 to capacitor C2 and the latter is connected to earth via R12. When the initial pulse 1 goes high at its end, Q1 becomes conductive and generates a positive pulse on line 102 as represented by the multiple Ie 6. Then C2 discharges through R11 and R12, the time constant being chosen so that a sufficiently long trigger pulse is available to actuate the second detectors stands. When the next initial pulse arrives, as in FIG

/ 23

- 23 -

shown in dashed lines, a negative differentiated impulse generated, but which is not effective for actuating the second detectors.

It is advantageous to only apply power to the probe coil 13 of FIG. 2 when measurements are actually carried out so that unnecessary heating is avoided. Accordingly, a switch 103 is provided, which takes the place 104 applies a positive voltage, as a result of which the transistors Q2 and Q3 are switched on via diodes D5 and D6. This prevents that the transistors produce the appropriate components of the composite output pulse. The switch 103 can if necessary be foot-operated.

During calibration and in position B of the puncture switch 77 of FIG. 6, continuous test pulses are applied from line 86 to transistor Q6 so that a continuous Series of test pulses generated by Q3 and the output lead 95 is fed. Q1 and Q2 are turned off by operating switch 105, so that on line 106 a positive Tension lies. This acts via the diode D7 and switches on Q1 and therefore Q2. The positive voltage on line 102 is also fed to line 102 via D8 and R13, keeping that line 102 high and the second detectors continuous holds open to let these quadrature signals through. In addition, the positive voltage on line 106 becomes the base supplied by Q10, which turns Q10 on and shorts the input of Q8. The regulating circuit is therefore ineffective and Q9 has its maximum conductivity given by R 14.

Figure 10 shows a CRT oscilloscope plot of the output of the 'M amplifier 62 (Figure 5) under normal operating conditions. The oscillocope is synchronized with the composite pulses. After an initial transition state at 111, the output remains!

• ■ · ¹ 48 I I I · • · * J t I : ι • · • I I 1 ■ * • * I
■ I 1 83Ό
! ■ II •

- 24 -

during the first half of the composite pulse (Fig. 4a) to approximately zero, as indicated at 112. At the start of the group of test pulses, the output rises to 113 and then falls to a stable value at 114. This follows another transition period 115 at the end of the pulse. As above explained, the second detectors let the outputs of the quadrature detectors at the end of the stable period 114 to the subsequent circuits.

The second detectors could be gated on a little earlier in the stable area 114. A different choice of duration the test pulse PRF and the start pulse can shorten the stable range, however, so that a triggering at its end is preferred. When the quadrature detectors hold their last output values, as is the case with read-out and hold detectors is the case, is a later triggering or keying of the second Detectors possible.

The areas 113 and 114 represent the output of the zero detector coils 18 and 18 'after amplification and filtering by the tuned circuit of the input amplifier 62. In Fig.10 is the oscilloscope gain is very high, so that very small imbalances in the zero detector coils are shown. When cracks occur, the amplitudes of the areas 113 and 114 change from group to group and the changes are detected by the following quadrature detectors.

In order to operate it is important to know that the tested material is saturated. For some applications it may be possible beforehand determine the current required for saturation. For example, if a sample of the pipe to be tested is available, can a crack of known size can be formed therein, such as a reference notch. This should be designed in this way be that any changes in the permeability of the Material can be avoided due to the formation of notches. When the pipe is then checked, it arises in the display circuits due to statistical changes in permeability a certain level of noise. The amplitude of the composite pulses can then be

- 25 - 'i! 148 830 /. . '' i ¹

• · Ϊ I rl

# ■ • 11 'till · 1

are increased until a notch signal of sufficient amplitude is obtained, which is, for example, twice the noise level. Having determined the required current level through this test different levels can be used in the actual test. Such a test is true for some Applications useful, but in practice it need not necessarily be sufficient because of the exact composition of the pipes to be tested is not known or an identical sample is not available.

The invention therefore provides a device for indicating whether the tested material is in the saturation range.

In Fig. 5, the output is a zero detection coil 18 through a C-R differentiating circuit 121 and then through an amplifier connected to an oscilloscope 123. If the material is in the Saturation area comes, occur in that of the detector coil 18

r- -, ,, iz or, distortions,,,.

recorded signal deformation products / caused by the Differentiating circuit are highlighted.

Fig. 11 indicates at 124 the type of waveform that is being used in the oscilloscope 123 can be observed. Due to the deformation products that occur when the material reaches the saturation range a signal 125 is generated. Line 126 corresponds to the beginning of the composite Irpufeas of Figure 4a and line 127 corresponds the end of the initial flat area and the start of the test pulses. For a given impulse current, the Signs or signals 125 in the case of light magnetizable materials earlier in the waveform, i.e. shifted towards left line 126. For materials that are difficult to saturate the display or signal 125 moves to the right. In the case of pipes made of the same material, however, they move in a very similar manner different wall thicknesses with a decrease in wall thickness the signals 125 to the left and with an increase in wall thickness to the right, Correspondingly, the characters or signals 125 can be used to select an operating current level that is sufficient Provides security that the material is actually driven to saturation.

/ 26

- 26 -

'..58 83ο .. ¹ ..

The area to the right of line 127 indicates the initial transition conditions during a group of test pulses, followed by a relatively stable level. This of course applies to a pickup or detector coil, so that the There is no zero effect of the other coil. ·

It is difficult to get the details of the signals 124, 125 with the state of the tested magnetic material in relation to one another, because changes in the normal and differential permeability the inductive resistance of the coils, which is influenced by changes in permeability, and the magnitudes of the DC voltage and AC components of the pulses, along with possibly other factors. Also is that It is difficult to state the exact influence of the differentiation, taking into account the preceding factors. At different Current levels and various probes and materials tested saw noticeable changes in the amplitude and shape of the wave established. Therefore, no adequate explanation for the results obtained is known at present. However, currently exists Reason to believe that the display will be useful in practice.

The preceding explanation of the specific embodiment relates primarily to the composite pulses according to FIG Fig.4a. If an operation according to Fig. 4b is desired, you can the pulse driver circuits of Fig. 8 can be modified accordingly, or other circuits can be used for this purpose to be developed. For example, Q3 can be switched differently, so that part of R3 is normally short-circuited, where the short circuit is canceled during the test pulse to thereby to change the partial action of R3 and R5. Or similarly, part of R5 can pass through Q3 during the occurrence of test pulses be short-circuited.

If an operation according to FIG. 4c is desired, the pulse generator according to FIG. 6 can be changed so that it has the test pulse group generated immediately after the initial pulse and the same resistors R3 and R4 in Fig. 8 can be selected. or

/ 27

- 27 -! I: 48 '8 "3Ct

■ III «>

it could be a composite group according to FIG. 4c in FIG. 6
are generated and supplied to each pulse input of FIG
will.

It is preferred to generate the composite pulses
4, as shown, to use DC voltage pulses. However, it is also possible instead of the one shown
Test pulse groups to use sinusoidal changes to the
AC voltage components for inducing eddy currents
to create in the object. Fig. 12 shows an example. Here is
the sine wave 131 is superimposed on the wide pulse 132. Also at
such composite pulses can use the quadrature detectors and the second detectors.

While it is preferred to work with composite pulses, simple DC pulse circuits may also be used for some applications, in particular for
Materials that are easy to saturate.

Fig. 13 shows the use of groups of pulses 133, 133 ',
which are separated by comparatively long time intervals 134.
As explained above, the DC voltage pulses have a DC voltage component, as indicated by the dashed line 135. With a duty cycle of 50%, the DC voltage level is half of the peak value. For pulses with
a pulse duty factor of 75%, the DC voltage level is higher,
at a duty cycle of 25% it is lower. Because the
Coil, to which the pulses are fed, an inductive resistor; a current builds up during each impulse and ': ■ drops something between the impulses, which creates a direct current. S.

In some cases the DC voltage or DC |

current level and the duration of a given pulse group to saturate ff

supply of the magnetic material is sufficient, whereby an ad-; t

satisfactory crack detection becomes possible. It can be with f-

Quadrature detector to be worked. A second ue-;.

tector keyed at or near the end of a pulse group ■

so that an indication is generated when the saturation is on (U

/ 28

- 28 -

is almost complete and the previous circuits have reached a comparatively stable state.

In some cases, with material that is easy to saturate, it may be possible to use a continuous series of DC voltage pulses to work without inadmissible heating. This is shown in FIG. 14. The duty cycle can be selected here so that a sufficient DC level is produced, as shown by line 136. Here too can work with quadrature detectors.

FIG. 15 shows a modified embodiment of the crack detector according to FIG. Only DC magnetization pulses are used here supplied from a pulse source 141 to the primary coil 13 of the probe according to FIG. The impulses can be like wave 4 8 and 9 can be generated. The recording or detector coils 18 and 18 'of the probe are here in a bridge circuit Center tap connected to the primary coil of transformer 142. The test signal from the test signal source 143, the one AC power source is fed to the bridge across a diagonal. The test signal can be in the form of intermittent groups of Pulses are present which and can be present during the last half of the respective magnetization pulses similar to FIGS. 4a and 4b in the manner of the wave 3 in FIGS. 8, 9 or can be generated directly after the saturation pulses are generated similar to Fig.4c. In the case of DC voltage test pulses, their DC voltage component helps maintain sufficient saturation of the inspected item, as discussed above. However, short groups can also be used sinusoidal waves similar to wave 131 in Fig. 12 can be used. Small amplitudes of the alternating voltage component can suffice and help to reduce the differential permeability of the article 10, as described above.

/ 29

- 29 -

The detector coils 18, 18 'are advantageously in series connected to one another or in a zero circuit, so that the voltages induced in them by means of the primary coil 13 cancel each other out. For the initial "leveling of the bridge." conventional trimming circuits can be used. The AC component of the test signal in the detector coils 18 and 18 'induce eddy currents in the object 10, their changes due to cracks etc. bring the bridge out of balance and a signal in the output winding 144 of the transformer. This signal is amplified and evaluated in quadrature detectors with a 90 ° phase shift second detectors triggered and then further processed and displayed, as described with reference to FIG. 5.

The gating in the second detectors 66 and 66 'should be opposite the beginnings of the DC magnetization pulses be delayed enough so that the resulting signals a adequate saturation of the object. If that Test signal only occurs during or immediately after a DC voltage pulse, the gating should also be compared to the subsequent beginnings of the test signal may be delayed so that transients are eliminated and output signals are created which correspond to stable values. With the detectors and the gating it is possible to have a continuous series Use short DC voltage pulses or a continuous sine wave as the test signal, provided that there is overheating is avoided because gating the influences of the test signal between DC magnetization pulses and turn off each DC magnetization pulse set during the initial period can, so that the resulting, detected signals a sufficient saturation of the Correspond to the subject. In general, the gassing takes place at intervals that are opposite to the associated beginnings of the DC magnetization pulses and the AC voltage components that occur during or immediately after the DC voltage pulses occur, are delayed.

/ 30

Here, too, a differentiation of the output signal can be used one of the detector coils 18 and 18 'are worked, which, as described with reference to FIGS. 5, 10 and 11, indicates saturation.

Fig. 16 shows a comparator for comparing a checked one Object 145 with a reference object 146. The tested object 145 is arranged or in a test coil arrangement is moved relative to it. The test coil assembly includes a Primary coil 147 and a secondary or take-up coil 151. A reference coil assembly comprising a primary coil 148 and a secondary coil 152 is provided for reference item 146. The coils can be inside or outside, depending on requirements the objects be arranged. Such a device is advantageous to determine changes in the dimensions, material, etc. of an object.

The primary coils 147 and 148 are assembled from a source 149 Signals with DC magnetization pulses and an AC component supplied, similar to that Crack detector according to Figure 5. The outputs of the secondary coils 151, 152 are advantageously against each other to form a zero circuit switched and fed to an amplifier, quadrature and second detectors and display circuits in block 153, which can be constructed similarly to FIG. Under control of the signals from 149, gating pulses for the quadrature and second detectors generated in 154 and applied to 153. If necessary, trimming circuits 155 can be used in a known manner will. The output of one of the coils 151 and 152 can also be differentiated, if necessary, to indicate the saturation. as described above.

With the help of quadrature detectors and gating signals which correspond to a sufficient saturation of the tested object to reduce the permeability to a small Decrease value, and can transition states: switched off so that signals correspond to stable values.

/ 31

- 31 - j I 48 830 ^: j "

■ ι «· * ι ι: t. ·

Fig. 17 shows a comparator with a bridge circuit similar of FIG. 15. Here, DC magnetization pulses from a source 141 are supplied to the primary coils 147 and 148. Test signals from a source 143 are fed to a diagonal of the bridge, which is provided by coils 151, 152 and resistors 161 and 162 is formed. The signal on the other diagonal of the bridge is transmitted via balancing circuits 163 (if necessary) Amplifier, quadrature evaluation, gating and display circuits 164 supplied. In pulse generators 65 and 67, as described above, on key signals for the quadrature and second Detectors are generated.

The resistor branches 161 and 162 are here in the bridge circuit and not shown in a center tap transformer; it can also connect other known bridge circuits, be turned.

Fig. 18 shows an arrangement in which an electrical balancing; _; : signal is generated and a reference object in a test coil | is used as a basis for a comparison with subsequent items to be tested. Here the coil 171 can be one of the f: detector coils 18, 18 'of the crack detector according to FIG. 5 or 15 or the test coil 151 of a comparator according to FIG. 16 or 17. The coil 171 is connected to a transformer 172 with a center tap via capacitors 173, which form a DC voltage insulation. If the circuit loads the coil 171 improperly, an isolation amplifier can be used.

Quadrature pulse pulses from the generator 174 / the taps of potentiometers 175 and 176, which are parallel to the primary coil of the transformer 172. The secondary coil of the transformer 172 is with Amplifier, quadrature detector, second detector and amplifier and display circuits similar to those described above.

The item 177 is first a reference object and the potentiometers 175 and 176 are set so that at 178 a reference ff is obtained. Then the reference object is removed and objects to be tested are used. differences

/ 32

■ · I · »ί

• ·

- 32 - 48 830

between a test object and the reference object are noticeable in deviations from the reference display at 178. Advantageously, the initial adjustment produces a balance that gives a zero reference indication so that small Deviations of the test items from the initial reference item and the direction of the deviation are displayed can.

Expectations:

/ 33

Claims (1)

. wtjESTHOPF PATENT Attorneys _for » _r ,,," ·· --BHfpjiiL.VKEDVvuESTHOFF (1917-1956) WUESTHOFF -ν. PECHMANN-BEHRENS-GOET2 dipl-incoerhard pum O ₉ J ₁ M ₉₇ ,) DIPL.-CHEM. DK. E. FREIHBRK BY FECHMANN FROFESSIONAL KEFRESENTATIVES BEFORE THB EUROPEAN PATENT OFFICB DR.-ING. DIETEK BEHRENS UANDATAiRES agrees pres l'office eurofebn des bkevbts dipl.-ing .; difl.-virtsch.-ino. rufbrtcobt D-8000 MUNICH 90 G 76 38 775.2 SCHWEIGERSTRASSE 2 Magnetic Analysis 1G-48 830 phone: (o8 ₉ ) itf * o51 telegram! protectpatent telex: j 24 070 December 12, 1980 Claim: Wirbelstroaiprüfgerät for testing pipes made of magnetic material for irregularities contained in them with a test head, the one from a DC voltage generator powered primary coil for generating magnetic flux of such magnitude in the article comprises that the object is magnetized up to its saturation region, where the permeability of the object is small compared to its maximum permeability, which is also fed by AC voltage signals Excitation coil arrangement for inducing eddy currents in the object and a detector coil device for Detecting changes in the eddy currents, with one with the output of the detector coil device connected detector device for generating signals which correspond to irregularities in the object, and with a display device responsive to the signals, characterized in that, that on the core (14) made of magnetic material of the test head (probe 11) which can be inserted into the tube (10) there is a surrounding it Primary coil (13) and on this a pair of zero coils of the detector coil device (18, 18 ') are attached. 1058