Device for controlling and measuring welding defects on a cylindrical wall and method implementing same

Abstract

A device and method for inspecting and measuring weld defects in a cylindrical wall of a cylindrical conduit. The device can include an inspection head forming a probe having a proximal end and a distal end along its longitudinal axis, and of which a first side called “inner side” is provided with at least one ultrasound wave transducer. The inspection head can include a second side, called “outer side” opposite the first side that has a curved surface in the form of a cylinder fraction, and wherein the curved surface of the second side has outward facing convexity. The wave transducer can be formed of a series of juxtaposed elements, each element being both a transmitter and receiver, wherein a surface of the series is curved and in the form of a cylinder fraction, and wherein the surface of the series has outward facing concavity.

Claims

1. A device for inspecting and measuring weld defects in a cylindrical wall of a cylindrical conduit, comprising: an inspection head forming a probe having a proximal end and a distal end along its longitudinal axis, and of which a first side called “inner side” is provided with at least one ultrasound wave transducer, wherein: the inspection head comprises a second side, called “outer side” opposite the first side that has a curved surface in the form of a cylinder fraction, with a second longitudinal axis parallel to the longitudinal axis of the inspection head, and wherein the curved surface of the second side has outward facing convexity; the wave transducer is formed of a series of juxtaposed elements, each element being both a transmitter and receiver, wherein a surface of the series is curved and in the form of a cylinder fraction having a third longitudinal axis that is co-planar with the longitudinal axis of the inspection head, and wherein the surface of the series has outward facing concavity; and a first plane parallel to the first side and a second plane parallel to the second side, wherein an angle between the first plane and the second plane is a nonzero acute angle β.

2. The device according to claim 1, wherein the angle β is approximately 21°.

3. The device according to claim 1, wherein the elements of the series are arranged one behind the other.

4. The device according to claim 1, wherein each element of the series is subdivided into several sub-elements distributed in a two-dimensional array, so that each sub-element has at least one neighbouring sub-element in a longitudinal direction and at least one neighbouring sub-element in a transverse direction.

5. The device according to claim 1, further comprising a manipulating pole having the inspection head at a distal end thereof.

6. The device according to claim 1, wherein a radius of curvature of the series of elements is between 8 and 30 mm.

7. The device according to claim 2, wherein the elements of the series are arranged one behind the other.

8. The device according to claim 6, wherein the radius of curvature of the series of elements is 10 mm.

9. A method for inspecting and measuring weld defects in a cylindrical wall of a cylindrical conduit, comprising: providing a device comprising an inspection head forming a probe having a proximal end and a distal end along its longitudinal axis, and of which a first side called “inner side” is provided with at least one ultrasound wave transducer, wherein: the inspection head comprises a second side, called “outer side” opposite the first side that has a curved surface in the form of a cylinder fraction, with a second longitudinal axis parallel to the longitudinal axis of the inspection head, and wherein the curved surface of the second side has outward facing convexity; the wave transducer is formed of a series of juxtaposed elements, each element being both a transmitter and receiver, wherein a surface of the series is curved and in the form of a cylinder fraction having a third longitudinal axis that is co-planar with the longitudinal axis of the inspection head, and wherein the surface of the series has outward facing concavity; and a first plane parallel to the first side and a second plane parallel to the second side, wherein an angle between the first plane and the second plane is a nonzero acute angle β; inserting the inspection head inside the cylindrical conduit, wherein the outer side of the inspection head has a same radius of curvature as an inner wall of the cylindrical conduit; applying the outer side against the inner wall; and scanning at least part of the inner wall during which the outer side remains in contact with the inner wall.

10. The method according to claim 9, further comprising: moving the inspection head in successive longitudinal movements inside the cylindrical conduit, and moving the inspection head in an angular movement between each longitudinal movement.

11. The method according to claim 9, further comprising: moving the inspection head in successive angular movements inside the cylindrical conduit, and moving the inspection head in a longitudinal movement between each angular movement.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Other characteristics and advantages of the invention will become apparent on reading the following description of one preferred embodiment of the invention. This description is given with reference to the appended drawings in which:

(2) FIG. 1 and FIG. 2, as mentioned above, illustrate the state of the art known to the present applicant;

(3) FIG. 3 is a simplified perspective view of the inspection head of a device conforming to the present invention in position within a VBP to be inspected;

(4) FIG. 4 is a perspective view of the inspection head shown in FIG. 3, here illustrated in a turned position at 180°;

(5) FIGS. 5 and 6 are respective front and side views of the inspection head in FIG. 4, in position within a VBP;

(6) FIG. 7 is a similar view to FIG. 6 of a variant of embodiment of the inspection head;

(7) FIG. 8 is an underside view of the head in FIG. 7;

(8) FIG. 9 is a schematic intended to illustrate how to implement the method of the invention;

(9) FIG. 10 is a schematic showing instrumentation for implementation of this method;

(10) FIGS. 11 to 14 are images showing the amplitude of the signals obtained when using a device of the invention.

DETAILED DESCRIPTION OF THE INVENTION

(11) The device 3 of the invention such as schematically illustrated in its entirety in FIG. 10 essentially comprises a pole 30 at the distal end of which there is an inspection head or probe 4.

(12) By convention, by distal end is meant the end the most distant from an operator positioned on the side of the pole opposite the head 4.

(13) A preferred example of embodiment of the head 4 is given in FIGS. 3 to 6.

(14) In these Figures, X-X′ is the longitudinal axis of the head and EP and ED are the proximal and distal ends respectively. In one nonillustrated embodiment these ends could be reversed.

(15) According to the invention, the inspection head 4, preferably made of electrically insulating material such as that known under the trade name “REXOLITE”, comprises a first side called “inner side” 42 and a second side 43 called “outer side” opposite the first side.

(16) This second side 43 has a curved surface in the form of a cylinder fraction having a longitudinal axis parallel to the longitudinal axis X-X′ of the head and with outward facing convexity. In FIG. 4, P1 is the plane containing the two end generatrixes of the cylinder fraction of the second side.

(17) The radius of curvature of this curved surface as will be seen below, is equal to the radius of curvature of the cylindrical conduit it is desired to inspect.

(18) The first side 42 only occupies a portion of the inner surface of the head and continues in the direction of the proximal end EP via a planar platform 40.

(19) The first side 42 receives an ultrasound wave transducer formed of a series of 5 juxtaposed transmitter/receiver elements 50, the series having a curved surface in the form of a cylinder fraction having the same direction as the longitudinal axis X-X′ and having outward facing concavity. Reference P2 is the plane containing the two end generatrixes of said cylinder fraction of this wave transducer. As an indication, the radius of curvature of the series 5 is between 8 and 30 mm and more preferably in the region of 10 mm.

(20) According to the invention, these two planes P1 and P2 together form a nonzero acute angle β.

(21) In particularly preferred manner, this angle is approximately 21°.

(22) In the embodiment shown in FIGS. 3 and 4, the elements 5 are arranged one behind the other with a narrow space between each one.

(23) However, and as shown in FIGS. 7 and 8, each element 50 can be subdivided into several sub-elements 501 distributed in a two-dimensional array, so that each sub-element 501 comprises at least one neighbouring subelement in a longitudinal direction and at least one neighbouring sub-element in a transverse direction.

(24) These views are merely intended to be illustrative. Therefore, for reasons of simplification the curvature of the elements 5 is not illustrated.

(25) At all events, each sub-element 501 is both a transmitter and a receiver.

(26) According to the embodiment shown in FIG. 4, at least eight aligned elements 50 are driven by electronics so as to excite the elements at particular determined instants as a function of the form it is desired to impart to the ultrasound beam generated by these elements.

(27) Angle β is about 21° to facilitate coverage of the angle ranges e.g. of the order of −20°, 70°. As will be seen below, the application of delayed excitation of the elements 50 already allows deflection of the beams at different angles, but in imperfect manner. With angle β it is possible to overcome this problem.

(28) The reference frame of the aforementioned angle range is the normal to the axis of the VBP tube, angle +90° being directed downwardly, with reference to FIG. 2.

(29) The dimensions of the head 4 (including the casing and cabling of the elements 5) must be smaller than the inner diameter of the VBP tube. Wave breakers, not illustrated, are advantageously positioned ahead (i.e. on the side of the proximal end EP) of the head 4 so that rebound echoes occurring inside the head do not mask the echoes sent by the defects to be inspected in the VBP weld.

(30) The operating frequency of the elements 50 is selected as a function of the depth (in relation to the sensor) of the defects it is desired to detect. For defects in the welded area of a VBP, a frequency of more or less 3 MHz is used.

(31) Inspection of defects of unknown orientation and position in the welded area is advantageously made possible by automation of inspection.

(32) For example, and as illustrated in FIG. 10, the device 3 of the invention comprises a pole 30 at the distal end of which there is the head 4 moved in self rotation by a motor 60, itself mounted on a rail 6, the latter sliding along a frame 61. It is thus possible to move the device 3 inside the VBP 14 both in the direction of the axis of the VBP 14 and in rotation about this same axis.

(33) In addition, for each position of the head 4, a series of ultrasound pulses is transmitted, each pulse allowing illumination of the medium in different directions. Therefore, contrary to the prior art, it is possible to detect defects of variable orientation having recourse to single set of elements 50.

Implementation—Example of Embodiment

(34) A piezoelectric transducer operating at a frequency of 3 MHz (bandwidth at 6 dB, 55% of the central frequency) with nine elements 50 is mounted on a head such as the one in FIG. 4, with an angle β of 21°.

(35) The elements 50 are of generally rectangular shape (dimensions 6 mm by 1.15 mm), aligned one behind the other, with a space of 0.15 mm between two elements. The head 4 is in Rexolite (registered trademark). Wave breakers not illustrated are positioned ahead of the head 4 i.e. on the side of its proximal end. The piezoelectric elements 50 are machined to impart a curved shape thereto with a radius of curvature of 10 mm.

(36) In this example of embodiment, the head 4 is integrated in a metallic casing of cylindrical shape facilitating guiding and holding thereof close against the inner wall 140 of the VBP 14.

(37) When set in operation, electronic apparatus of “M2M” trade mark is used to excite transmission by the piezoelectric elements 50 and to record the signals received by these same elements 50. The electronics are guided by “multi2000” software.

(38) This software computes time delays to be applied to the excitations of the different elements 50 of the transducer, as a function of the shape it is desired to impart to the emitted ultrasound beam. In this application, the time delays are computed so that the ultrasound beam of longitudinal waves is deflected by an angle of between −20° et 70°, with a pitch of 2°.

(39) The head 4 is moved along the axis of the VBP tube 14 (so-called ‘scanning’ axis as indicated by the arrow f in FIG. 9). Signals are transmitted and recorded every 0.5 millimetres.

(40) Measurement is repeated after successive rotations of the head 4 about the axis of the tube with a pitch of 2° (arrow g in FIG. 9). This makes it possible to cover the entire welded area to be inspected.

(41) Movement of the probe is ensured by a motorised arm of “Micro-contrôle” trade mark driven by a controller of same trade mark (reference ITL09).

(42) The measured signals are then post-processed using “CIVA” software. This allows images to be obtained representing the amplitude of the received signals as per a colour code, as a function either of the position of the head 4 along the scanning axis, or of the position of the head 4 according to the rotation value or as a function of the applied ultrasound pulse (i.e. as a function of the deflection angle of the ultrasound beam). Analysis of these images allows deducing of the presence of defects in the VBP 14.

(43) This procedure is illustrated below.

(44) A model is considered representing a welded VBP in every aspect comprising two artificial defects (obtained by electro-erosion) called ‘circumferential’ defects in the welded area.

(45) These are defects called EE1 and EE2. The data are analysed using processing software (here “CIVA” software) which allows data to be represented in the form of different types of images.

(46) In particular so-called ‘C-scan’ images are seen which represent the amplitude received by the transducer in colour code (variations in grey shade in FIG. 11 et seq.) as a function—vertically—of the position of the head 4 in rotation, and—horizontally—as a function of the scanning axis.

(47) Post-processing then involves observing the images of ‘C-scan’ type for different angles of sonication and determining which sonication angle allows the obtaining of an indication of maximum amplitude received from a defect.

(48) More specifically, the following steps are carried out using CIVA software (however any other software can be used to process these data):

(49) For all sequences and all sonications, selection of a time window excluding so-called ‘permanent’ echoes due to reflections in the wedge;

(50) Also, selection of the range of scanning values (along the vertical axis) of the area corresponding to the weld (omission of the area corresponding to the tube alone via observation of rebound echoes on the tube with a sonication at 0° (i.e. perpendicular to the axis of revolution of the tube);

(51) Next, observation of the C-scan type image (maximum amplitude over time at each scanning pitch and increment, for one determined sonication), and changes as a function of the chosen sonication. When an echo appears in the welded area in the C-scan, the shape of this echo can be seen in the B-scan;

(52) The increment is then selected corresponding to the observed echo, and the B-scan corresponding to this sonication and to this increment is displayed;

(53) To maximise the signal-to-noise ratio, it is possible to fine-tune the choice of angle of sonication by selecting the previously obtained scanning position and increment and then observing the so-called S-scan which represents the signal measured as a function of time and as a function of sonication angle for the fixed scanning and increment;

(54) The choice of time limits, scanning limits and even range of sonication angles can also be based on CIVA simulation. For a set of potential defects in the welded area, this allows computing of the arrival times of the echoes from these defects for each angle of sonication, and an order of magnitude of the amplitude thereof. This can be used to support analysis of experimental signals.

(55) It can also be mentioned that parasitic echoes causing most of the noise correspond to geometric echoes (rebound off the bottom of the tube). Therefore, simulation (or measurement on a gauge block) first allows identification of the type of these echoes on blind inspection of a new component of same geometry, and secondly image processing (not tested here) of wavelet filtering type would allow suppression of these echoes to the benefit solely of defect echoes.

(56) This data analysis shows that the defect EE1 is detected when the beam is deflected by 12°, and defect EE2 is detected with a deflection of 26°. The images below show the ‘C scan’ mage obtained and the so-called ‘B-scan’ image, the latter representing the amplitude of the signal as a function of time (vertical axis) and of the position of the head 4 along the scanning axis (horizontal axis), the rotation position being selected so that the signal from the defect is of maximum amplitude.

(57) Defects EE1 and EE2 are therefore very well detected by means of the presence of an echo (dark spot in the images) clearly distinguished from the other ‘noise’ signals.

(58) Therefore, by implementing the method just described that proceeds by scanning, preferably automated, the entire surface to be inspected and by applying “multi-pulses” i.e. ultrasound emissions spaced over time, optimal quality of defect detection is obtained even when the number of transducer elements 50 is reduced. This particularly applies when they are nine in number as illustrated in the Figures.

(59) With adapted processing software such as the one cited above, echoes can be evidenced which would have been lost within the noise of the structure without visualisation of the spatial dimension of the measured echoes.