Multiple coil eddy current probe and method of flaw detection.

Technical Field

*Inconel is a registered U.S. trademark belonging to the International Nickel Company, Inc.

This invention generally relates to an eddy current probe system and method, and is specifically concerned with an improved probe system having a plurality of concentrically-disposed coils for simultaneously detect­ing different types of flaws or discontinuities at differ­ent depths within a wall of, for example, an Inconel* tube.

BACKGROUND OF THE INVENTION

Eddy current probes for inspecting the walls of metallic conduits are known in the prior art. Such probes are particularly useful in inspecting the Inconel tubes used as heat exchangers in nuclear steam generators for flaws caused by corrosion or fretting. Generally, these eddy current probes comprise a coil mounted in a probe head that is slidably movable within the interior of the tube being inspected, and electrically connected to a current generator which conducts an alternating current through the coil as it is moved. The current generator is typically capable of generating alternating currents having frequen­cies of between 10 kHz and 1 MHz. An impedance detecting circuit, which may take the form of an inductive bridge, is also connected across the leads of the coil. In operation, the alternating current conducted through the coil excites it into generating a pulsating magnetic field whose magni­tude and polarity changes in accordance with the frequency of the current. When the coil of the probe is positioned in the vicinity of an electrically conductive wall, the changing magnetic flux emanating from the coil induces eddy currents in a portion of the wall. The particular voltage, amperage and direction of the eddy currents produced are dependent in part upon the specific impedance of the portion of the wall that conducts the eddy current. Because the direction of flow of the eddy currents generat­ed by the coil is opposite to the current flowing through the probe coil, the magnetic field created by the eddy currents creates an impedance in the probe coil. The impedance experienced by the eddy currents is in turn dependent upon the resistance these currents encounter as they circulate through the wall. Since flaws in the metal wall (such as cracks, pits, or regions of local thinning) create regions of higher resistances at the flaw locations, eddy current probes may be used to locate flaws by con­stantly monitoring the impedances of the coils as the probe coils are moved along the walls of the tube. Sharp changes in impedance over localized areas would indicate the existence of cracks or pits or other relatively small area flaws, whereas gradual changes in impedance over a broad region of the conduit might indicate large-area flaws such as a grain change in the metal, an area of material creep, or a thinned wall region.

Typically, such prior art eddy current probes utilize either a single bobbin type coil whose axis of rotation is parallel with the longitudinal axis of the conduit being inspected (and which is operated in an "absolute" mode), or a pair of bobbin coils having the same radius which are spaced apart from one another (which are operated in a "differential" mode). In either case, the probe head containing either the single or the double coil is moved within the interior of the tube along its longitu­dinal axis. More recently, "pancake-type" probe coil configurations have come in to use. In such configura­tions, the axis of rotation of the windings of a relatively flat coil is disposed along the radius of the conduit, and the coil is used to scan the inner wall of the conduit by moving it both radially and longitudinally, thereby impart­ing a helical motion to the coil. Such pancake-type coils are capable of more accurately locating the precise point where some types of flaws reside in the conduit wall. Both configurations have been successfully used in hostile environments where direct inspection of tubing is impossi­ble, such as the walls of the approximately 40 miles of Inconel tubes used as heat exchangers in nuclear steam generators.

While such probes are capable of performing satisfactory inspections of such heat exchanger tubes, the applicants have noted a number of problems associated with these probes which, up to now, has limited their useful­ness. For example, since the depth of penetration of a particular pulsating magnetic field is dependent upon its frequency, it is difficult or impossible for a coil con­ducting an alternating current of a fixed frequency to simultaneously and reliably resolve flaws at all depths of the conduit wall. Secondly, while a small-diametered coil is best able to accurately pinpoint the location of a particular flaw, such a coil i incapable of transmitting pulsating magnetic fields having frequencies low enough to completely penetrate the conduit walls if the coil diameter is made too small. The applicants have observed that these constraints often necessitate multiple scans of the tube wall with different diametered coils operating at different frequencies if all of the flaws therein are to be accurate­ly and reliably located. However, the multiple scanning of a particular tube wall with different probe coils operating at different current frequencies greatly protracts the time necessary for the testing which in turn results in increased down-time for the steam generator being inspect­ed. As the typical revenue losses associated with such generator down-time often exceeds $100,000.00 a day, the expenses associated with the necessary multiple scans are very substantial. Moreover, such multiple scanning also increases the time that the inspecting personnel are exposed to potentially hazardous radiation, which adds even more to the cost of the eddy current testing.

In view of the foregoing, the applicants have concluded that there is a need for an eddy current probe system which is capable of accurately and reliably plotting all types of flaws at all different depths within the walls of a small diametered metallic tube with only a single rapid scan. Ideally, such an eddy current probe system would provide better resolution than any prior art probe system or combination of any such systems. Finally, the eddy current probe system should be extremely versatile, and capable of instantaneously adjusting its pulsating magnetic fields for maximum coupling with the flaw areas detected so as to afford maximum resolution under a broad range of conditions.

DISCLOSURE OF THE INVENTION

Generally speaking, the invention is an improved eddy current probe for simultaneously detecting different types of flaws at different depths. The improved probe includes a current generator for generating alternating currents of substantially different frequencies, and a plurality of concentrically arranged coils, each of which separately communicates with the current generating means and each of which conducts an alternating current of substantially different frequency for detecting different types of flaws at different depths within the metallic wall. A magnetic shielding material is disposed between the edges of adjacent coils for preventing cross-talk between each coil and the pulsating magnetic fields gener­ated by adjacent coils. The probe system further includes an impedance detector circuit for providing an electrical output representative of the relative impedance of the respective coils. Both the current generator means and impedance detector circuit may be parts of a commercially available eddy current testing device.

Preferably, the highest frequency alternating current is conducted through the coil having the smallest diameter, whereas the lowest frequency alternating current is conducted through the coil having the largest diameter, for two reasons. First, the application of a relatively low frequency alternating current to the largest diameter coil insures that at least one of the pulsating fields emanating from the coils will penetrate far enough into the tube wall and resolve flaws located on the exterior tube surface, such as areas of grain change, creep and wall thinning due to corrosion, fretting or mechanical scratch­ing. Second, the application of a relatively high frequen­cy alternating current to the smallest diameter coil not only allows this coil to detect and to pinpoint the loca­tion of small area flaws such as cracks and pits, but also creates a small and well defined reference point that the probe system operator may use to infer the location of all of the other coils in the probe head at all times. The probe head of the preferred embodiment includes three concentrically disposed coils of progressively larger diameter which may conduct frequencies as high as 1.0 MHz and as low as 30 kHz. The inner coil preferably initially conducts an alternating current of approximately 500 kHz, the middle coil a current of approximately 250 kHz, and the outer coil a current of approximately 50 kHz.

Finally, the improved eddy current probe system may include a computer connected to both the impedance detector circuit and the current generator for adjusting and multiplexing the frequencies of the alternating current conductedhrough the coils of the probe head as the head is moved along a selected portion of the metallic wall and comparing the resulting amounts of coil impedance detected in order to minimize the impedance of each coil and hence to maximize the flaw-detecting resolution of the probe head.

In the method of the invention, the computer receives and stores the value of the impedance of each coil as the probe is helically moved around the interior of the tube wall. When impedance changes are detected, the computer incrementally adjusts the frequencies of these currents either upwardly or downwardly, and compares the values of the impedances of each coil associated with both the old frequency and the new incrementally adjusted frequency. The computer makes subsequent incremental adjustments to the alternating current frequencies based on the resulting comparisons for the purpose of arriving at the minimum impedance values achievable with the coils. When such maximum impedances are obtained, the resolution of the probe head, as a whole, is maximized.

The probe system of the invention obviates the need for multiple scans of the tube wall being inspected. The concentric arrangement of the coils locates flaws and other discontinuities in a faster and more accurate manner than would a colinear or other multiple arrangement of such coils due to the clear reference point provided by the centrally disposed, small diameter coil which preferably transmits a sharply focused, high-frequency pulsating magnetic field. Finally, the near-instantaneous, impedance minimizing adjustments of the frequencies of the alternat­ing currents conducted through the various diametered coils insured that the resolution of the probe head will be maximized at every point of inspection without any signifi­cant sacrifice in scanning time.

BRIEF DESCRIPTION OF THE SEVERAL FIGURES

<ul id="ul0001" list-style="none"><li>Figure 1 is a schematic representation of the overall system of the invention, illustrating both the manner in which the probe head scans a tube, the manner in which the probe output is displayed, and the lay-out of the components of the eddy current circuitry that implements the method of the invention;</li><li>Figure 2 is an exploded, side view of the probe head of the invention;</li><li>Figure 3A is a plan view of the probe head illustrated in Figure 2 with the coils shown in phantom;</li><li>Figure 3B is a cross-sectional side view of the probe head illustrated in Figure 3A;</li><li>Figures 4A and 4B are a schematic diagram of a pancake-type eddy current probe coil, and a graph repre­senting the relationship between the depth of penetration of the pulsating magnetic field emanated by a particular probe coil as a function both of the diameter of the probe, and the frequency of the alternating current applied thereto, and</li><li>Figures 5A and 5B together form a flow chart illustrating the preferred method of the invention as implemented by the circuitry illustrated in Figure 1.</li></ul>

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

General Description of the Structure and Method of the Invention

With reference now to Figure 1, wherein like numbers designate like components throughout all the several figures, the eddy current probe system 1 of the invention is particularly well adapted for inspecting the walls 2 of an Inconel tube 3 of the type used in nuclear steam generators for different types of flaws at different depths within the walls 2, such as cracks 4 (both axially and radially aligned), pits 5, and broader sections 6 of the tube wall which have been subjected to creep or grain change, or thinned by corrosion, fretting or mechanical abrasion. The probe system 1 includes a probe head 8 which emanates a magnetic field 9 generated by a multiplicity of concentrically-arranged coils, each of which conducts a different frequency alternating current. The probe head 8 is mechanically mounted in a probe body (not shown) that is moved along a helical path by a drive unit 8.5 that in­cludes an electric drive motor coupled to an optical encoder. The drive unit 8.5 is selectively actuated by a switching circuit 9.5 that is in turn connected to a power source (not shown).

The probe head 8 is electrically connected via a cable to eddy current circuitry generally designated at 10. This eddy current circuitry 10 includes a computer-compatible frequency multiplexer 11 which, in the preferred embodiment, is a model MIZ 18 manufactured by ZETEC, Inc., located in Isaquah, Washington, USA. The frequency multiplexer 11 is connected to a computer 12 which is programmed to adjust the frequencies of the currents conducted through the coils of the probe head 8 to optimize resolution whenever an impedance change indicative of a flaw or other discontinuity is detected. In the preferred embodiment, the computer 12 is preferably a Model HP 9836 and an IEEE 488 interface circuit manufactured by Hewlett-Packard located in Palo Alto, California, USA, modified by a Zetec DDA-4 processing package.

The output of the computer 12 is connected to a cathode ray tube (CRT) 13 which displays, through selected computer graphics, the impedance changes experienced by the coils within the probe head 8 as they are helically moved around the inner surface of the wall 2 of the Inconel tube 3. The frequency multiplexer 11 includes a current genera­tor 14 for generating a multiplicity of alternating cur­rents. The leads of the concentrically arranged coils within the probe head 8 are connected to the current generator 14 by way of a switching circuit 15. This switching circuit 15 acts to selectively conduct the same current through one or more of the concentrically arranged coils in the probe head 8 when the overall resolution of the system 1 would be enhanced by such an interconnection. The current generator 14 is connected to and controlled by a frequency adjusting and timing circuit 16. The circuit 16 is capable of not only changing the frequency of the currents conducted through each of the concentrically arranged coils of the probe head 8, but also of multiplex­ing a sequence of different-frequency currents through each of the coils. Finally, the frequency multiplexer 11 includes an impedance measuring circuit 17 which may be an inductive bridge connected to the current generator 14 for measuring changes in impedance experienced by the concen­trically arranged coils in the probe head 8 as they are helically moved along the tube 3. *Delrin is a registered U.S. trademark belonging to E.I. DuPont De Nemours and Company of Wilmington, Delaware, USA.

In operation, the probe head 8 is mounted within a probe body, such as the Delrin* probe carrier of the combined ultrasonic and eddy current inspection probe described and claimed in co-pending U.S. Patent Application Serial No. 079,860, filed July 30, 1987, by Thomas Arzenti and assigned to the Westinghouse Electric Corporation, the entire specification of which is incorporated herein by reference. The probe carrier of the aforementioned probe helically moves the probe head 8 in wiping engagement with the inner surface of the wall 2 of an Inconel tube 3 to scan the same. As is indicated in Figure 1, such helical motion includes an axial component along the longitudinal axis Y of the tube 3, as well as an angular component w along the radium r of the tube 3. As will be described in more detail hereinafter, as the probe head 8 is scanningly moved throughout the inner surface of the tube wall 2, different frequencies of alternating current are conducted through different probe coils so as to generate a plurality of different shaped pulsating magnetic fields. At least one of these coils generates a field that penetrates completely through the wall 2 of the tube 3 so that exteri­or surface flaws such as the thinned portion 6 of the tube may be detected. At least one of the coils generates a magnetic field which interacts substantially with the inner surface of the wall 2 of the tube 3 so that interior surface flaws such as pit 5 which exist predominantly in the interior portion of the wall 2. Finally, at least one of the coils in the probe head 8 generates a pulsating

magnetic field whose lines of flux interact principally with the interior portion of the wall 2 in order to resolve flaws such as the cracks 4 which extends completely through the wall 2 of the tube 3.

The results of the scan may be displayed on the CRT tube 13 by way of computer graphics wherein the verti­cal axis designates the impedance changes experienced by each of the coils, the perspective axis represents the longitudinal axis Y of the tube 3 and the horizontal axis represents the angular location of the probe head 8 with respect to the radius of the tube 3. Preferably, each one of the ten scans illustrated on the CRT screen 13 includes a different-colored output for each of the coils in the probe head 8. In the preferred embodiment of the inven­tion, three coils are used in the probe head 8, although many more of such coils may be used if desired. In the graph displayed on CRT screen 13, the output of the small-frequency coil is illustrated by a small-dashed line, while the output of the medium and low frequency coils are represented by large-dashed and solid lines, respectively. In the preferred embodiment, the small, medium and large diametered coils are initially operated at 500, 250 and 50 kHz respectively, and are adjusted to maximize the resolu­tion of the probe system 1 when an impedance change indica­tive of a flaw is detected.

Specific Description of the Structure and Method of the Invention

*Co-netic and Netic are registered U.S. trademarks owned by the Magnetic Shield Division of Perfection Mica Co. of Bensenville, Illinois, USA.

With reference now to Figures 2, 3A and 3B, the exterior of the probe head 8 is formed by a casing 19 preferably fabricated from a plastic material such as Delrin. The casing 19 includes a ring-shaped side wall 20 for retaining the coils and the probe head 8, as well as a circular back plate 21 having a centrally disposed mounting lug 22 projecting therefrom. In the preferred embodiment, the ring-shaped side wall 20, the circular back plate 21 and mounting lug 22 are all integrally formed. The interi­or of the probe head 8 includes outer, middle and inner shield walls 23, 25 and 27, as well as outer, middle and inner probe coils 32, 34, 36. Each of the shield walls 23, 25 and 27 is formed from a ring 29 of material that is very high in magnetic permeability which is closely circum­scribed by a ring 30 of material which is somewhat lower in magnetic permeability but which has higher saturation characteristics than the ring 29 that it circumscribes. In the preferred embodiment, the inner ring 29 of each shield is formed from Netic* foil. The ring-shaped shield walls 23, 25 and 27 insulate each of the coils 32, 34 and 36 from the pulsating magnetic fields created by the coils adjacent thereto by providing a low reluctance path that diverts potentially interfering magnetic fields from the coil surrounded thereby. Such shielding in turn prevents inter coil cross-talk during the operation of the probe system 1. As is evident in Figure 2, each of the coils 32, 34 and 36 includes a pair of leads 33a, 33b, 35a, 35b and 37a, 37b which are ultimately connected to the frequency multiplexer 11 by way of a cable as indicated in Figure 1. Also, a ferrite focusing core 39 is centrally disposed within the inner coil 36 in order to both focus and intensify the pulsating magnetic field generated thereby, thereby advan­tageously creating a reference point for inferring the location of the surrounding coils 32 and 34. While a total of three coils are shown in the preferred embodiment, a greater number of coils could be used if desired.

In the preferred embodiment, both the outer, middle and inner probe coils 32, 34 and 36 are formed from about thirty windings of copper wire. As the outer diame ter of the probe head 8 is only about 4.7625 millimeters, these windings are formed from very thin wire. In the alternative, the coil windings may also be formed from state-of-the-art printed circuit board techniques. Probe head coils formed by way of this latter method are advanta­geously flatter than coils formed by bobbin-like windings of copper wire. Such extreme flatness might have the affect of increasing the overall resolution of the probe head 8 since the fields generated by the coils 32, 34 and 36 would not significantly project out of the back side of the probe head 8, where they might have an opportunity to interact with a metallic body which is not the particular portion of the conduit wall 2 being inspected. In both embodiments, the front face of the probe head 8 is coated with a thin layer of Delrin or some other self-lubricating plastic material both to protect the delicate coil wind­ings, and to minimize friction as the probe head 8 helical­ly rides along the inner surface of the tube wall 2 being inspected.

With reference now to Figures 3A and 3B, the casing 19 of the probe head 8 further includes three pairs of lead bores 41a, 41b, 42a, 42b, and 43a, 43b for conduct­ing the lead wires 33a, 33b, 35a, 35b, and 37a, 37b of the coils 32, 34 and 36 respectively. As is indicated in Figure 1, these lead wires are ultimately connected to the switching circuit 15 of the frequency multiplexer 11 by way of a cable which extends through the drive unit 8.5 by way of appropriate slip-ring connectors (not shown).

The general operation of the probe head 8 may best be understood with reference to Figures 3B, 4A and 4B. At the beginning of each data increment, the frequencies of the currents conducted through the coils 32, 34 and 36 are selected so that the field 45 emanated by the largest-diametered coil 32 will penetrate completely through the wall 2 of the Inconel tube 3 being inspected, while the field 47 emanated by the smallest-diametered coil 36 will only shallowly penetrate into the inner surface of the wall 2. The field 46 emanated by the middle-diametered coil 34 interacts primarily with the interior of the wall 2 of the tube 3. Such a field shape insures that no signifi­ cant discontinuity at any point long the wall thickness will be overlooked by the probe system 1.

While it would be possible to operate the largest-diametered coil 32 at the highest frequency, and the smallest diametered coil 36 at the lowest frequency, such a mode of operation is normally not preferred by virtue of the limitations illustrated in the graph of Figure 4B. This graph demonstrates that the depth of penetration of the pulsating magnetic field emanating from a particular coil is dependent not only upon the frequency of the current conducted through the coil, but also upon the diameter of the coil. This graph also illustrates that, in order to achieve a particular depth of penetration P, the coil diameter D must be of a certain minimal size before any effective penetration can take place, no matter how low the frequency of the current conducted thereby. Hence, the diameter of the smallest-diametered coil 36 may well be below the minimum size diameter necessary to project a pulsating magnetic field completely through the wall 2 of an Inconel tube 3. Of course, the diameter of the smallest diametered coil 36 could be made large enough to project a field through the wall 2. But if this were done, the resolution ability of the probe head 8 might be seriously diminished, since the resulting large coils would not be capable of accurately locating the borders of small area flaws. For all these reasons, the smallest-diametered coil 36 is generally operated at the highest frequency, while the largest-diametered coil 32 is operated at the lowest frequencies.

Figures 5A and 5B form a flow chart which specif­ically represents the method of the invention as imple­mented by the computer 12. After the probe body (not shown) carrying the probe head 8 has been slidably inserted into a particular heat exchanger tube 3 so that the probe head 8 is adjacent to the region of the tube 3 that the operator wishes to inspect, the operator first selects the precise axial length of the tube 3 that he wishes to scan as indicated by box 52. Next, the computer 12 converts this axial length into numbers of scans, as is indicated by box 54. This is a simple operation, which is determined on the basis of the pitch of the thread of the screw (not shown) used to impart helical motion to the probe body.

After the computer 12 has determined the number of scans corresponding to the selected axial length of the tube 3 to be inspected, the system operator next selects an angular increment of data acquisition, as is indicated by box 56. If the operator desires a relatively quick, coarse resolution scan of the tube 3, this angular increment may be as high as 15 degrees. If, on the other hand, he wishes to have a relatively quick, fine resolution scan made of the axial length being inspected, this angular increment may be as low as 1 degree. Normally, to provide for a uniform resolution throughout the entire 360 degrees of the scan, the angular increment chosen will be a number of degrees which is evenly divisible into 360 degrees.

In the next step of the method, the computer 12 determines the total number of data increments Nt to be taken by determining the number of angular increments in each scan, and multiplying times the number of scans computed in step 54. After this has been accomplished, the drive unit 8.5 that moves the probe head 8 is actuated as is indicated by step 60. This is accomplished by the computer 12 actuating the power switching circuit 9.5, which connects the electric motor of the drive unit 8.5 to a source of electrical power (not shown). Immediately after the motor has been actuated, the output of the optical encoder coupled to the drive unit motor is fed into the input of the computer 12.

Once the computer 12 begins to receive an output from the optical encoder of the drive unit 8.5, it inguires whether or not the encoder rotates past a zero degree position within a selected time period, as is indicated by question block 62. If the computer 12 does not receive such a confirmation signal, it displays an error condition, as is indicated by block 63, which informs the operator of the probe system 1 that the motor and encoding unit 8.5 is not properly rotating the probe head 8. If, on the other hand, it receives a signal that indicates that the encoder of the motor and encoder unit 8.5 has swept past a zero degree position within a selected time period, it proceeds to block 64, and actuates the eddy current circuitry 10.

After the eddy current circuitry 10 has been actuated, the computer 12 proceeds to question block 66, and inquires whether or not the drive unit 8.5 has moved the probe head 8 to the beginning of the first data incre­ment. As the data increments are only about 0.152 millime­ters apart when the probe head 8 is used to inspect a 15.875 millimeters diameter heat exchanger tube at 1-degree increments, the amount of time necessary for the computer 12 to answer the inquiry in question block 66 in the affirmative is normally only a small fraction of a second. As soon as the computer 12 detects that the beginning of the first data increment has arrived, it immediately begins storing the impedances associated with each coil 32, 34 and 36 as indicated in box 68. At substantially the same time, the computer 12 instructs the frequency multiplexer 11 to conduct current of initial frequencies F1, F2 and F3 through the coils 32, 34 and 36. In the case of an Inconel heat exchanger tube 3 having a wall thickness of approxi­mately F1, F2 and F3 will correspond to 50 kHz, 250 kHz, and 500 kHz, respectively when the outer diameter of the coil 32 is approximately 4.7625 millimeters and the outer diameter of the inner coil 36 is approximately 1.5875 millimeters. The applicants have observed that the conduc­tion of currents of such frequencies into coils of such diameters result in magnetic fields that interact with tube walls of such thicknesses in much the fashion shown in Figure 3B, wherein the field lines 45 emanated by the largest-diametered coil 32 penetrate completely through the wall 2 of the tube 3, and the field lines 46 and 47 of the mid-diametered and smallest-diametered coils 34 and 36 penetrate the interior and the inner surface of the wall 2 as illustrated.

Immediately at the commencement of the first data increment, the computer proceeds to question block 72, and inquires whether or not the eddy current circuitry 10 detects any impedance changes in any of the coils from millisecond to millisecond. If the answer to this inquiry is affirmative (as would be the case when the probe head 8 is swept adjacent to a flaw in the tube wall 2) the comput­er 12 implements the step indicated in block, 74, and immediately begins changing the frequencies of the currents conducted through the coils 32, 34 and 36 in order to optimize resolution. Every time a current of a new fre­quency is conducted through one of the coils, the computer proceeds to question block 76 and inquires whether or not the impedance experienced by any of the coils 32, 34 and 36 has decreased as a result of the changes in frequency of the current conducted therethrough. If the answer to this question is affirmative, then the computer 12 returns to the step indicated in block 74, and continues to change the frequency of the current on a millisecond to millisecond basis. Of course, the practical effect of the steps indicated in blocks 74 and 76 is for the computer 12 to search out and find the frequency of the currents in the coils 32, 34, and 36 which minimize the coil impedance, and thereby maximize the resolution of the flaw in the tube wall 2.

Once these optimum frequencies have been found, the answer to the inquiry in question block 76 will become negative, and the computer 12 will proceed to further optimize resolution by optimizing the phase angle of the alternating currents flowing through the coils 32, 34 and 36, and then optimizing the amplitudes of the resulting signals as is indicated in blocks 78 and 80.

Immediately after all of the optimization steps 74 through 80 have been completed, the computer will proceed to question block 84, and ask whether or not the next data increment N+1 has arrived. If the answer is negative, then the computer 12 will re-execute the step indicated in block 70, and again conduct the initially chosen current frequencies through the various coils, which again will likely result in a near repeat of the steps indicated in block 72 through 80. However, when the answer to this inquiry is affirmative, the computer computes the cylindrical coordinates associated with the data increment step N+1 as is indicated in block 86. It will then immedi­ately display the minimum impedance value associated with each of the coils 32, 34 and 36 within the increment as is indicated in block 88. Lowered impedance values will have the effect of generating hills in the graphic display on the CRT tube 13, such as illustrated in scans 2 through 5 in Figure 1.

Next, as is indicated by question block 90, the computer 12 will inquire whether or not the data increment just completed the total number of data increments Nt. If the answer to this question is negative, the computer 12 recommences steps 70 through 90. When, however, the answer to this inquiry is affirmative, the computer 12 stops the operation of the probe system 1, as is indicated by block 92.

Of course, if the computer 12 detects no discon­tinuities in the tube wall 2 in a particular data incre­ment, it answers the question in question block 72 in the negative, and continues to conduct currents through the coils 32, 34 and 36 at the initially chosen frequencies until the end of that data increment. Following this, the computer 12 displays the impedance values at the cylindri­cal coordinates associated with the no-flaw increment. As the impedance values remain unchanged throughout such a data increment, the resulting display is a straight seg­ment. If no impedance changes are detected throughout a complete scan, the resulting display is a straight line for the entire scan (such as scan No. 1. in Figure 1).

While the flowchart of the method of the invention has been described only in terms of changing the frequencies of the currents conducted through the coils 32, 34 and 36 in order to achieve the desired result of maximum resolution, the general program may be embellished with one or more subroutines which would enhance the resolution of the probe head 8 even more. For example, when the smallest-diametered coil 36 detects a flaw on the inner surface of the tube wall 2, the concentrically disposed middle and outer coils 34 and 36 could be operated at the same high frequency for a very small fraction of a second. If one or more of these larger diametered coils simultane­ously interacted with this inner surface flaw, the result­ing information could help determine the area of the flaw since the area of the coils in the probe head 8 is known. Still another subroutine that might be added to the general program is the operation of two of the three coils in a differential mode, in contrast to the operation of all three coils in an absolute mode.

While the aforementioned method of the invention has been described with respect to the use of three coils, it is also adaptable for use with a probe head 8 having more than three coils. If more than three coils are used, a larger span if initially chosen frequencies are prefera­bly used, which would have the effect of increasing the overall resolution of the system 1 even more.<tables id="tabl0001" num="0001"><img file="EP0332048A2_D0001.tif" /></tables><tables id="tabl0002" num="0002"><img file="EP0332048A2_D0002.tif" /></tables>