WO1992012422A1

WO1992012422A1 - A.c. field measurement testing system

Info

Publication number: WO1992012422A1
Application number: PCT/GB1992/000053
Authority: WO
Inventors: David Anthony Topp; Martin Christopher Lugg
Original assignee: Technical Software Consultants Limited
Priority date: 1991-01-11
Filing date: 1992-01-10
Publication date: 1992-07-23
Also published as: US5574376A; CA2099785A1; NO932466D0; EP0566624B1; NO308562B1; DE69211532D1; DE69211532T2; NO932466L; CA2099785C; GB9100589D0; EP0566624A1

Abstract

An a.c. field measurement testing system comprises a probe including a yoke (10) for inducing a uniform a.c. field, sensor means (7, 8, 9) comprising a first coil (8) having a longitudinal axis extending in a first direction and a second coil (9) having a longitudinal axis extending in a second direction orthogonal to the first direction and means (11) for effecting energisation of the field, the arrangement being such that signals produced by the coils (8, 9) indicative of defects in a test piece are processed whereby one component of a magnetic field is plotted against another in a display.

Description

A.C. FIELD MEASUREMENT TESTING SYSTEM

This invention relates to an a.c. field measurement testing system.

The techniques currently in use for the detection and sizing of surface-breaking cracks on ferritic steel structures require access to the bright metal. It is therefore necessary to carry out extensive pre-cleaning of the area to be inspected. This is particularly onerous underwater where marine growth and calcareous deposits have to be removed as well as oxide layers prior to inspection. Any protective paint layer present also has to be removed and then reapplied after inspection. Magnetic* particle inspection (MPI) is then commonly used for crack detection and crack length measurement, possibly followed by a.c. potential drop (ACPD) inspection if crack depth information is required but more often material is removed by grinding until the MPI indication disappears. The a.c. field measurement (ACFM) technique was developed to extend the crack sizing capability of ACPD to a non-contacting form capable of inspecting through the non¬ conducting coatings. ACFM therefore differs from ACPD only by its measurement of above surface magnetic fields instead of surface potential differences. As long as the surface current density induced into the specimen is locally of uniform strength and direction, theoretical modelling of the expected magnetic fields above a defect is possible which allows crack sizes to be estimated from experimental measurements. However, a problem associated with ACFM is that signals indicative of defects in a test piece are conditioned by the speed at which a probe is scanned relative to the test piece and it is often difficult to distinguish between cracks in a test piece and spurious signals resulting from probe lift off.

According to the present invention there is provided an a.c. field measurement testing system comprising a probe and a processor including at least one display, the probe including a yoke for inducing a uniform a.c. field in a first direction, sensor means comprising a first coil having a longitudinal axis extending in the said first direction and a second coil having a longitudinal axis extending in a second direction orthogonal to the first direction and means for effecting energisation of the field the arrangement being such that signals produced by the coils indicative of defects in a test piece, when the probe is located adjacent the test piece, are processed by the processor and one component of magnetic field is plotted against another in the display.

The present invention is designed to implement a practical application of the ACFM technique by a combination of novel probe and graphical display formats. The probe is designed to achieve the two requirements of ACFM by the production of a locally uniform input current, perpendicular to the expected crack edge, and by the simultaneous measurement of at least two mutually perpendicular components of magnetic field at the same point in space. The technique can also be implemented with an array of double coil measurements, to give instantaneous display of the surface magnetic field distribution, or with controlled probe motion to give improved spatial resolution. The remote graphical display combines the conventional time-base plots of the individual magnetic field component strengths with a novel display format in which one component is used as the abscissa and another as the ordinate.

This format, which has been dubbed a "butterfly plot" because of the characteristic shape produced by a crack, has several practical advantages for crack detection. Firstly, the removal of the time-base gives a display independent of the speed at which the probe is scanned and which can be retraced for confirmation. Secondly, the simultaneous monitoring of two components helps to distinguish between cracks and spurious indications due to probe lift-off. The combination of point measurements of magnetic field components and

"butterfly plot" format allow multidimensional threshold boxes to be used to simplify or automate the detection process. It is also highly advantageous to be able to store the traces and replay them at a later date.

The ferromagnetic yoke is used to induce a current into the test piece while at the same time minimising the amount of direct induction into the sense coils. This is necessary to allow comparison of experimental readings with those theoretically predicted, and is achieved because the bulk of the magnetic field is channelled thorough the pole pieces onto the test piece underneath which completes the magnetic path. A current sheet is thereby induced in the test peice in a direction normal to the line joining the pole pieces, the current density being approximately uniform over a large percentage of the area directly under the yoke. The sensing elements are sited midway between the pole pieces close to the test piece in a region of optimum field uniformity and consist of at least two coils wound one on top of another centred on a single point in space. These coils may be "x-oriented", i.e. with an axis parallel to the line joining the pole pieces, "y-oriented", i.e. with an axis perpendicular to the line joining the pole pieces and parallel to the surface of the test piece, or "z-oriented", i.e. with an axis normal to the surface of the test piece. Theoretical modelling has shown that an x-oriented coil measures the background field strength which is reduced in the presence of a surface breaking defect going the whole length of the defect, a z-oriented coil measures a positive-going peak (or crest) in field strength above one crack end and a negative-going peak (or trough) above the other end, and a y-oriented coil measures a signal similar to a z-oriented coil but which reverses sign on traversing the plane of the crack. The coils may be of a circular or polygonal cross-section with dimensions small enough to allow good spatial resolution of the magnetic fields but large enough to enable local variations in geometry or lift-off to be smoothed out. Experience has found that diameters between 2mm and 10mm are desirable. The coils may be wound in either absolute or differential mode, the latter with two halves of the windings would be in opposite senses. The spacing between the poles of the yoke should be large enough to allow a reasonable area of uniform field strength but small enough to cope with curvature and other geometry effects in the test piece. Again, experience has found that spacing of between 25mm and 75mm is ideal. The height of the ends of the pole pieces above the base of the probe should be small enough to minimise direct induction into the sense coils but not so small that small changes in lift-off cause large variations in input field strength. Experience has found that a height between 5mm and 20mm is ideal.

The invention will now be further described, by way of example only, and with reference to the accompanying drawings in which:

Figure 1 is a schematic view of one embodiment of an a.c. field measurement testing system in accordance with the invention; Figure 2a is an end view of a probe head of the system; Figure 2b is a detailed view of a sensing element located in the probe head; Figure 2c is a side view of an alternative probe head containing an array of sensing elements; Figure 3 is a representation of a graphical display of the system showing examples of timebase and

"butterfly" plots arising from a surface breaking defect.

Referring to Figure 1 , the ACFM testing system of the present invention makes use of an ACFM Crack Miσrogauge 2. This may be housed in a watertight enclosure for subsea operation. This unit outputs the drive current to the yoke in probe head 1 and amplifies, filters and rectifies the signals from the sense coils in the probe head 1. The signals are then transmitted via a serial communication line in an umbilical cable 4 to a computer 3.

The probe head 1 consists of a housing approximately 50mm wide by 25mm deep by 25mm high to which is attached a handle approximately 75mm long. Referring now to Figure 2a which is an end view of the layout of the probe head, the sensing element is housed in a cylindrical dimple 5 protruding from a base. Two stainless steel legs, 6, also protrude from the base to permit the probe to accommodate a range of curvatures in the test pieces. The sense element is located as close to a front of the probe head as possible to allow it to be located as close to a weld toe as possible.

The sensing element is shown in Figure 2b and comprises two absolute coils wound around a common central point, 7. One coil, 8, is x-oriented, while the other coil, 9, is z- oriented. The x-oriented coil consists of 50 turns of 0.07mm diameter enamel coated copper wire and has a square cross- section of side 5mm. The z-oriented coil consists of 100 turns of the same wire and has a circular cross-section of diameter 3mm. The excitation yoke, 10, is shown in Figure 2a and comprises a thin flat mild steel plate with two ends bent through 90 degrees into a plane orthogonal to the base of the probe. Several turns of 1mm diameter PVC coated copper wire, 11, carrying a 1.0 A current at 5kHz are wound around the mid section of the yoke.

A stainless steel plate, 12, 0.5mm thick, is fixed to the base of the probe to minimise wear. The thickness of this plate is small compared to the skin depth in stainless steel at 5kHz, so that the magnetic fields passing through the sense coils are essentially unaffected.

The electronic instrumentation consists of a twin channel input amplifier module in which the voltages induced in the two coils are sampled alternately at a rate of 100 Hz via a solid state relay. This rate is sufficient to give almost simultaneous monitoring of the two channels. The signals are filtered and passed through a phase-sensitive detector which is preset to the phase of the background field measured by the x-oriented coil away from a defect. The instrument contains a microprocessor which controls sensor selection, gain selection, analog to digital conversion of the signals and communication with the computer via a serial link. The in-air version of the instrument also contains a simple keypad and display for control without the need for a computer. The underwater version of the instrument contains no external controls, but has a simple graphical display on a flying lead.

A menu-driven software program is used to control the acquisition, display, storage and interpretation of readings taken from the instrument. The user-friendly program incorporates automatic scaling of graphical displays, control of sampling rates and threshold values and storage of data in a form compatible with a standard spreadsheet packag including a facility for attaching half a screen o descriptive text. Chosen parameter settings can be stored an recalled as required, and previously recorded data can b replayed.

Referring now to Figure 3, which is a representation of one form of graphical display on the VDU of the voltages plotted as they are received from the system described above, the left hand side of the screen, 13, contains plots of signal strength versus time for each of the two sense coils contained in the probe head, while the right hand side, 14, simultaneously records the signal strength from the x-oriented coil versus the signal strength from the z-oriented coil (i.e. the "butterfly" plot) . The recordings shown in Figure 3 represent typical displays resulting when the probe head is traversed along a simple surface-breaking crack. On the three traces shown, points 15 represent the start of each trace, points 16 mark the start of the crack, points 17 mark the midpoint of the crack, points 18 mark the end of the crack, and points 19 mark the end of each trace. In order for a crack to be differentiated from other variations in signal arising from probe lift-off or local geometry changes, threshold limits are required on the signals from both sense coils simultaneously. This is conveniently done in the "butterfly" plot, 14, by the addition of a two-dimensional threshold area, 20. Variations of signal that remain inside this area are ignored, but excursions of the trace outside this area result in the triggering of a warning that is transmitted to the operator by both audible and visible means. Confirmation of such a feature may be achieved by reversing the direction of probe movement which will result in the retracing of the "butterfly" plot in the opposite sense. The retracing is possible with a hand-held probe head because of the shape and size of the "butterfly". This retracing plot is independent of the speed of probe movement.

A second example of a possible application of the invention is shown in Figure 2c which depicts a side view of an alternative probe head containing a one-dimensional array of sense elements for inspection of thread roots. The probe head has a thread form on one face to mate with the thread to be inspected. Each crown of this thread contains a pair of sense coils 21 located at a point as described previously. The ferromagnetic yoke, 22, has a dimension of 50mm between pole pieces and a length sufficient to cover all the sense elements.

The electronic instrumentation required for operation of the probe is enlarged from that previously described to include a multi-channel input multiplexer with a corresponding increase in the complexity of controlling software and graphical display.

Claims

CLAIMS 1. An a.c. field measurement testing system comprising a probe (1) and a processor including at least one display (3) characterised in that the probe (1) includes a yoke (10) for inducing a uniform a.c. field in a first direction, sensor means (7, 8, 9) comprising a first coil (8) having a longitudinal axis extending in the said first direction and a second coil (9) having a longitudinal axis extending in a second direction orthogonal to the first direction and means (11) for effecting energisation of the field, the arrangement being such that signals produced by the coils (8, 9) indicative of defects in a test piece, when the probe (1) is located adjacent the test piece, are processed by the processor and one component of magnetic field is plotted against another in the display (3).

2. A system as in Claim 1 wherein the yoke (10) has a dimension from one pole to another of between 25mm and 75mm.

3. A system as in Claim 1 or Claim 2 wherein each of the coils (8, 9) is of a diameter of between 1mm and 10mm.

4. A system as claimed in any one of the preceding Claims wherein a surface (12) of the probe (1) which is intended to be located adjacent the test piece is of non- ferromagnetic material.

5. A system as claimed in any one of the preceding Claims, wherein the yoke (10) comprises a rectilinear portion and pole pieces extending at right angles to the rectilinear portion.

6. A system as claimed in any one of the preceding Claims wherein the sensor means (7, 8, 9) is provided with a third coil a central longitudinal axis of which is orthogonal to the axes of the first (8) and second coils (9).

7. A system as claimed in any one of the preceding Claims wherein there is provided a plurality of sensor means (22).

8. A system as claimed in Claim 7 wherein the sensor means (22) are connected such that, in operation of the system, the sensor means (22) are sampled sequentially.