Inspection of header tube holes.

This invention relates to the inspection of header tube hole surfaces and, more particularly, to scanning apparatus for and methods of inspecting surfaces of header tube holes.

Headers are used extensively in steam boilers as a means for joining fluid circuits and for distributing fluid to fluid circuits. Such headers typically comprise a large-diameter, heavy-walled cylindrical shell, circular in cross-section, having multiple straight or bent tube stubs which extend partially into holes formed through the wall of the header. The tube stubs are fixed to the header via rolled tube joints or by welded tube connections. The end of the portion of each tube stub extending outside of the header is designed to be connected to a component of a fluid circuit, usually in the form of a tube which is welded to the tube stub.

Cracks that initiate from the header surface surrounding the tube holes in steam headers of boilers in fossil fuel burning electrical power plants can lead to failures which can cause costly unscheduled plant outages. If cracks are detected in early stages, however, plans can be made to repair or replace the header during scheduled maintenance and repair outages. Detection and characterization of cracks near the surface surrounding holes in structural components, by nondestructive testing, is often a key factor in assessing the condition and remaining useful life of a component. Crack characterization, i.e. measurement or estimates of depth, length and location of cracks that initiate from the surface of holes in headers, is used with additional information to predict remaining life and to plan repair or replacement of the headers.

Cracks have been detected with boroscopes that use fibre optics technology. Access to the hole is obtained by cutting and removing a section of the tube or tube stub that leads into the header. The boroscope is manually inserted into the hole and rotated to inspect the entire circumference. However, proper focusing of the boroscope is difficult to achieve due to instrument sensitivity to changes in probe-to-surface distance. Therefore, quality and reliability of data is less than adequate without a mechanism which permits a controlled scan that maintains a relatively constant position and alignment of the probe. The same, as well as other difficulties, arise with regard to the use of other sensors including ultrasonic transducers and eddy current probes.

In accordance with a preferred embodiment of the apparatus of this invention, an inspection probe or the like is mounted to a tube stub mounting mechanism, connected to a drive means which rotates and translates the probe within the header tube hole. The mounting mechanism is interconnected between the probe assembly and the drive means. The mounting mechanism is designed to be detachably connected to the tube stub for supporting the apparatus. The probe includes spring means for pressing a probe sensor into engagement with the surface of the header surrounding the header tube hole.

In the preferred embodiment, the tube mounting mechanism includes a central hollow cylinder connected to the spring means. The spring means, and hence the sensor, is movable responsive to movement of the hollow cylinder. A flanged tube is slidably disposed on the hollow cylinder. The tube mounting mechanism further includes a pair of expansible rings on the tubes and means on the tube for expanding the rings into supporting engagement with the tube stub. The hollow cylinder may be constructed of a flexible material for scanning bent tube stubs. The drive means includes means for rotating and axially translating the hollow cylinder and the sensor therewith.

The preferred apparatus is used for supporting and positioning an inspection probe or the like within a header tube hole. It can be used for examining successive portions of the surface of the header surrounding the header tube hole by means of the inspection probe or the like. The apparatus can carry out a method of inspecting header tube holes by performing rotary or axial scans, or both, to detect cracks and to calculate the depth of the cracks. An eddy current inspection technique is preferably utilized to obtain data for such calculations.

The invention will now be further described, by way of illustrative and non-limiting example, with reference to the accompanying drawings, in which like reference numerals designate like or corresponding parts throughout, and in which: <ul id="ul0001" list-style="none"><li>Figure 1 is a perspective view, partly broken away, of a portion of a header;</li><li>Figure 2 is a partial sectional view of a header and a tube received within a header tube hole;</li><li>Figure 3 is an exploded side view, partly in section, of an apparatus according to a preferred embodiment of the invention;</li><li>Figure 4 is a schematic illustration of an apparatus like that illustrated in Figure 3 mounted in position within a header tube hole for inspecting the header hole in accordance with a method embodying the invention; and</li><li>Figure 5 is a side view of a tube stub mounting mechanism in accordance with another embodiment of the invention.</li></ul>

Referring now to the drawings in detail and in particular to Fig. 1, there is shown a cylindrical header 20 which has a plurality of holes 22 extending through the thickness of the header wall from the inner surface 16 to the outer surface 18 of the header 20. A plurality of tube stubs 24 are received within the holes 22 and are fixed to the header 20 by welding, as illustrated by welds 26 in Fig. 2. Alternatively, the tube stubs 24 may be fixed to the header by formation of a rolled joint. An end 28 of each tube stub 24 is located within the tube hole 22 at a position intermediate the inner and outer surfaces 16, 18 of the header 20 and distant from the inner surface 16 of the header 20.

The opposite ends 29 of each tube stub 24, which are located outside of the header 20, are designed to be connected to tubes (not shown) of an external fluid circuit to provide fluid communication between the header 20 and the fluid circuit for joining fluid streams passed to the header 20 or for distributing fluid streams passed from the header 20.

On occasion, it becomes necessary to disconnect the tubes of the fluid circuit from the tube stubs 24 and to inspect the peripheral surface 14 surrounding the header tube hole 22 and located intermediate the inner surface 16 and the end 28 of the tube stub 24 which is seated within the wall of the header 20.

Referring to Fig. 3, an inspection apparatus, according to a preferred embodiment of the invention, includes a probe assembly 30 which is capable of being inserted through the tube stubs 24 of the header 20 and into the header tube holes 22. The probe assembly 30 is mounted to a tube stub mounting mechanism 40 which is designed to engage one of the tube stubs 24. Both the probe assembly 30 and the tube stub mounting mechanism 40 are connected to a drive carriage 50 which is operable to move the probe assembly 30 as is more particularly des­cribed hereafter.

The probe assembly 30 is provided with sensors 31, 32 supported for movement on one end of substantially parallel spring bars 33, 34, respectively. Each spring bar 33, 34 at its opposite end is secured, in cantilevered fashion, by a support plate, referred to herein as clamp block 35 composed of a number of plates which are bolted together. The clamp block 35, which is located at the leading end of the probe assembly 30, in turn, is mounted to an elongated support bar 36 which extends, away from the leading end of the probe assembly 30, between and substantially parallel to the spring bars 33, 34 and terminates beyond the spring bars 33, 34 in a C-shaped stop plate. The stop plate is composed of a crossbar 37, which extends substantially perpendicularly relative to the elongated axis of the support bar 36, and stop lugs 38, 39 at opposite ends of the crossbar 37 which overlap the ends of the respective spring bars 33, 34. An internally-­threaded, tubular socket 27 is provided on the side of the stop plate opposite the support bar 36.

The tube stub mounting mechanism 40 is connected to the probe assembly 30, in the illustrated embodiment, by threaded engagement of a threaded end of a hollow cyl­inder 42 and the socket 27. As shown in Fig. 3, a flanged tube is mounted about the hollow cylinder 42. The hollow cylinder 42 may be moved freely relative to the flanged tube. The flanged tube has a flange 25 located at an end near the end of the hollow cylinder 42 which is coupled to the socket 27. Resilient locking rings or collars 43, 47, a sleeve 45, and spacer washers 44, 46, 48 are slidably received upon the tube 41 of the flanged tube intermediate the flange 25 of the flanged tube and an adjusting knob 49. The adjusting knob 49 is threadably engaged to the tube 41. The resilient locking collars 43, 47, in the preferred embodiment, have a diameter slightly larger than the sleeve 45, flange 25 and spacer washers 44,46, 48 and are designed to approximate the internal diameter of the tube stubs 24.

The drive carriage 50 comprises a pair of gener­ally parallel mounting plates 51, 52 with guide rods 53, 54 which extend parallel to each other.

The end of the flanged tube opposite the flange 25 is engaged to mounting plate 51 provided as part of the drive carriage 50. The hollow cylinder 42 freely passes through mounting plate 51 and is operatively mounted to mounting plate 52 in a manner more particularly described hereafter.

The hollow cylinder 42 is operatively attached to a gear 55 for rotation via a bearing mounted on the mount­ing plate 52. The gear 55 is connected to a rotational drive motor 60 or other means for imparting rotation to the hollow cylinder 42 via an intermediate gear 56.

Mounting plate 52 is slidably disposed on the guide rods 53, 54 and has a pair of bearings 57, 58 which are slidable along the guide rods 53, 54, respectively, to allow the mounting plate 52, with the hollow cylinder 42, gears 55, 56 and motor 60, to move rectilinearly and paral­lel to the guide rods 53, 54.

A pinion gear 61 drives a drive rack 62 to recti­linearly advance and retract the mounting plate 52. The drive rack 62 is connected to mounting plate 52 and freely extends through mounting plate 51. The pinion gear 61 is mounted on a drive shaft 63 which may be driven by suitable means, for example, a hand crank or stepping motor. In addition, a pair of pinion gears 61 may be provided on opposite sides of the carriage assembly, interconnected by drive shaft 63, for engagement with parallel drive racks.

Figure 4 is a schematic representation showing the above-described apparatus in position for inspecting the header tube hole 14 of header 20.

In operation, the tube stub 24, which is seated within the hole which is to be inspected, is disconnected from the fluid circuit to which it is normally attached. The probe assembly 30 is inserted through the tube stub 24 and into the hole 22 until the tube stub supporting mechan­ism 40 is positioned with the tube stub 24. As the probe assembly 30 passes through the tube stub 24, the spring bars 33, 34,which resiliently urge the sensors 31, 32 outwardly of the support bar 36 are compressed towards the support bar 36. On passing tube stub end 28, however, the spring bars 33, 34 expand outwardly to press the sensors 31, 32 against the surface 22. The outward movement of the spring bars is restricted by contact with the stop lugs 38, 39 of the C-shaped stop plate. The stop lugs 38, 39 restrict lateral movement of the spring bars 33, 34 beyond the inner diameter of the tube stub 24 so that the probe assembly can be retrieved from the header 20 without sustaining damage to the probe and sensors. The sensors 31, 32,in the preferred embodiment, comprise housings having tapered edges to further facilitate retrieval of the probe assembly 30.

The adjusting knob 49 is then tightened. The tightening of the adjusting knob 49 causes displacement of the sleeve 45, and spacer washers 44, 46, 48 towards the flange 25. This causes the resilient, expansible rings or locking collars 43, 47 to diametrically expand into sup­porting engagement with the tube stub 24. The circumferen­tial contact of resilient locking collars which may, for example, be rubber rings, with the tube stub 24, facil­itates support and reference for the apparatus.

As illustrated in Figs. 3 and 4, the hollow cyl­inder 42 comprises a rigid tubular member composed, for example, of type 304 stainless steel.

In an alternative embodiment, the hollow cylinder 42 may be a flexible tubular member, for example, a braided steel tube. Braided steel flexible tubes are typically flexible to bending but extremely rigid to torsional forc­es. Utilization of a flexible hollow cylinder will permit the probe to be inserted through bent tube stubs. In such case, the tube stub mounting mechanism can be modified, as shown in Fig. 5, to omit the central spacer washers, and alternatively to include two short sleeves 65, 66.

The electrical connections to the probe, e.g., electrical lines 21, 23, are contained within the hollow cylinder 42. The gears 55, 56 mounted at the end of the cylinder 42 link the probe assembly 30 to the rotational drive motor 60. The cylinder 42 rotates in the bearing 59 within the mounting plate 52. A second motor 64, illus­trated in Fig. 4, is operable to drive the pinion gear 61 and drive rack 62 to slide the mounting plate 52 along the guide rods 53, 54 so that the probe assembly 30 moves in the axial direction of the header tube hole. The pinion gear 61 is preferably mounted to the mounting plate 51. Control can be provided for both motors 60, 64 for a coordinated sequence of rotational probe scans at various axial positions along the length of the header tube hole. When stepping motors are used, relatively precise rota­ tional and axial positioning of the probe assembly sensors 31, 32 can be obtained.

Although two springs 33, 34, as illustrated in Figure 4, are provided to support two support sensors 31, 32, there are many applications where only one sensor is needed. In such case, a housing fixture that does not contain a sensor is fastened to one spring, to provide a counter force to balance the other sensor spring. This counter force helps keep the probe assembly 30 near the central axis of the hole. The forces of the leaf springs tend to move the sensor into engagement with the surface of the hole. The sensor housing can be mounted on sponge rubber pads to assist in alignment of the sensors. The flat spring and sensor backing, in such case, provide sufficient surface area for application of an adhesive bond to secure the sponge rubber pad.

The above-described apparatus provides a system­atic controlled means for scanning one or more transducers, or other inspection devices, over the surface of a header tube hole. It allows good transducer alignment with rela­tively constant probe-to-tube surface distance throughout the scan. Moreover, the use of electronically controlled stepping motors provides precise rotational and axial posi­tioning of the sensors. The use of separate motors for sensor rotation and axial movement offers particular vers­atility in scanning the surface of the header hole. Rota­tional scanning can be executed without slip rings and without twisting the sensor cables. This can be accom­plished by programming the stepping motor control to return to starting position before repeating the scan at a new axial location. Stepping motors, moreover, can be associa­ted with digital control to facilitate data acquisition at predesignated locations for repeated scans. This provides spacially coherent signal responses which are the basis for improved accuracy in the interpretation of sensor data.

The above-described apparatus has a simple design which can be embodied in a light weight structure for ease of use. The expandable rubber rings hold the device in place throughout the inspection. Finally, the stop lugs of the probe assembly and the tapered sensor housings facilitate the removal of the device from a tube hole without damage to the probe, cabling or spring mounts.

The above-described apparatus has particular ut­ility in the sensing and processing of eddy current signal response data to provide on estimation of the depths of cracks that initiate from the peripheral surface of header tube holes. An eddy current probe is preferably embodied within the one or more sensors 31, 32 of the probe assembly 30. The eddy current probe may be among one of several commercially available type probes and is operatively connected, via conventional electric lines 21, 23 or cable extending from the sensor housing, through the hollow cylinder 42.

The eddy current sensor 31, 32 is scanned along a circumferential path on the surface 14 of the hole 22. Eddy currents are induced in the electrically conducting header material by applying alternating currents to the eddy current coils. The probe is designed so that the components of the induced eddy currents are in a direction that is perpendicular to the plane of the crack. The well known skin effect associated with the eddy currently pheno­menon concentrates the eddy currents near the surface of the header material. The induced currents tend to flow along the crack surfaces around the bottom and ends of the crack. This change in current path results in a corres­ponding change in at the alternative magnetic field. The change in magnetic field in turn results in a change in the amplitude and phase of the voltage observed at the term­inals of the coil. Commercially available eddy current instruments detect the alternative voltage to provide a signal response as the probe is scanned over a crack.

Eddy current signal responses caused by variables such as electrical conductivity, magnetic permeability and small variations in probe-to-material distance can be reduced or minimized by the incorporation of two coils. The coils can be positioned and oriented so that one coil exhibits a substantially different response to the crack than the other. For example, cross-wound coils can be designed and oriented so that one coil produces a minimum signal response to the crack while the other produces a maximum response. Since the coils are connected so that their respective signals subtract the signals caused by nondirectional variables such as electrical conductivity, magnetic permeability and probe-to-material distance will tend to cancel. Other differential coil configurations such as concentric coils and adjacent coils of various shapes and sizes can be used to achieve a similar reduction of these unwanted signal responses.

The eddy current probe can also be designed so that a relatively high frequency will provide greater res­olution than that of a lower frequency. A multifrequency eddy current instrument can be used to obtain separate responses to two or more adjacent cracks. Appropriate signal responses have been obtained, for example, by using a cross-wound coil that is shielded with a thin brass foil approximately 0.127mm (0.005 in) thick. The brass foil contains a small hole at the position where coil windings cross. Rel­atively low frequencies, e.g. 1-10 KHz, penetrate through the foil. With this low frequency excitation, the coil provides a good response to increasing crack depth but can be influenced by two or more adjacent cracks at a given time. This multiple crack condition causes a signal response of greater width and amplitude than would occur for a single crack having identical dimensions. Rela­tively high frequencies, e.g., 500 KHz, produce a magnetic field that penetrates primarily through the small hole in the brass shield. This provides a high resolution indica­tion and precise location of each of the cracks.

Low frequency reponse to multiple cracks is depen­dent on the number, spacing and lengths of the cracks as well as their respective depths. For example, two or more adjacent cracks can cause depth indications that are twice that of only one crack. The use of additional higher fre­quencies provides a basis for compensating for the errors in measuring crack depth when multiple cracks occur.

In accordance with a preferred technique for det­ecting and measuring cracks initiating from the surface 14 surrounding a header tube hole 22, the eddy current sen­sors, energized via the electrical lines 21, 23, to induce eddy currents in the surface to be inspected, are scanned in a circumferential direction by rotating the hollow cylinder 42. The scan may be repeated at additional axial locations by axially moving the hollow cylinder via opera­tion of the drive rack 62. Signal responses are obtained at each location. In particular, two signal responses, such as the in-phase component and quadurature component or amplitude and phase is recorded for each frequency.

Various other features may be extracted from the signals and recorded in respect of the axial position of the scan.

The crack depth is then computed by using a formula that compensates for the errors caused by the proximity of two or more adjacent cracks of variable length and depth.

An example of a formula for estimating crack depth is given as follows:<maths id="math0001" num=""><img file="EP0242947A2_D0001.tif" /></maths>

In general, the formula is a function of variables as indicated by: D = F(A_km, A_ki, d_i, l_j) where, A_km = Maximum low-frequency signal component, k, acquired within a set of rotational scans A_ki = Signal amplitude of one of the k components at circumferential location X_i X_i = Circumferential location of crack, i, determined from the high-frequency, high-resolution signal response Z_ij = Axial locaton of the end j of crack i d_i = Estimated lateral distance of crack at location X_i, from the position X_m, where the maximum occurs, i.e., (X_m - X_i) l_j = Estimated length of an adjacent crack at lateral position j, e.g., (Z_it - Z_ib) where t represents the top end of crack i and b represent the bottom end.

C_km and C_ki for i = 1, 2, ... N are coefficients. The designation k = 1 indicates that the amplitudes are acquired from the first component of the low frequency signal reponse.

In the case where there are no adjacent cracks, the formula reduces to the simple form: D = C_m(A_lm)²

Similar formulae incorporating additional extracted features such as crack length and phase angle can be de­rived. Although the feature extraction and depth calcula­tions can be determined by manual computation, microcompu­ter software can be designed to perform these computations with much greater efficiency and reliability. The weighting coefficients, i.e., C_km and C_kj, can be determined by using the least squared error criterion and guided random search techniques can be used to estimate coefficients when sufficient test sample data are available.

The technique generates and uses multifrequency eddy current signal response data to provide improved estimates of crack characteristics such as crack depth.

A microcomputer to facilitate scan control, extraction of the numerical value of features and esti­mation of crack characteristics, e.g., depth and length. The preferred eddy current technique permits detection of cracks that may be missed by other inspection techniques such as penetrant and magnetic particle inspections. It does not require fluid coupling that may contaminate comp­onents and is relatively fast compared to other inspection methods and associated devices.