Field of the Invention The present invention concerns a method and apparatus for inspecting rails. More particularly, but not

exclusively, this invention concerns a method and apparatus for investigating the propagation of cracks in a railway rail. The invention also concerns a method and apparatus for calibrating a rail inspection system comprising an eddy current, sensor.

Background of the Invention Metals cracking under conditions of repeated stress cycling is a well known phenomenon. Eddy current sensors may be used to inspect a metal article for cracks. For example, UK Patent Publication Number GB 2469009 discloses an eddy current sensor that may be used in the railway industry for inspecting rails for cracks. . In the majority of cases, a metal article is stressed by repeated pure bending. As a consequence, cracks grow into the material at right angles to the outer surface of the article. However, railway rails are stressed in an extremely complex manner with the result that cracks propagate through the rail at a wide variety of angles to the surface of the rail, for example, between 5 degrees and 25 degrees to the surface, as opposed to 90 degrees as is more conventional. There is no way of predicting what direction a crack will propagate in due to the complexity of the loading as described.

Therefore, while eddy current sensors may detect the presence of a crack in a rail, it may not be possible to accurately determine how the crack has propagated through the rail. This may cause problems when attempting to classify the seriousness of a crack, for example when determining whether or not the section of rail needs replacing. While eddy current systems may be calibrated using a series of notches in a rail test piece, each cut to a different depth by electrical discharge machining, these notches are cut at 90 degrees to the surface of the rail. As mentioned, crack propagation in a rail is often at angles other than 90 degrees to the surface of the rail, and so such calibration may not be accurate.

Conventional crack propagation investigation techniques may comprise the vertical sectioning of a rail, polishing the cut surface, and then taking an image of the polished surface. However, such a technique is time consuming, difficult to control accurately, and expensive.

Additionally, only a single crack may be investigated at a time .

The present invention seeks to mitigate the above- mentioned problems.

Summary of the Invention

The present invention provides, according to a first aspect, a method of investigating the propagation of a crack in a rail, the method comprising the following steps:

recording an image of the top surface of the rail, removing a layer of material from the top surface of the rail;

repeating the steps of recording an image of the top surface of the rail and removing a layer of material from the top surface of the rail a plurality of times; and analysing the plurality of images obtained to track the propagation of the crack through the rail.

The cross section of a rail typically provides an asymmetric I-beam, comprising a head, a web, and a foot. The top surface of the rail should be understood to mean the exposed top surface of the head of the rail. Typically this will be the uppermost portion of the rail when the rail has been installed for use, but top surface should not be considered to. limit the invention to a rail being disposed in any particular plane. Therefore, while typically the method of inspection will include the rail being disposed with the top surface uppermost, it also covers the rail being disposed on a side during the crack investigation process, so the top surface is in an approximately vertical plane.

The layers of material may be removed in a plane approximately parallel to the plane of the top surface of the rail. Where the top surface of the rail is curved, the layers of material may be removed in a plane approximately tangential to the middle point of the top surface. The layers of material may be removed in an ^' approximately horizontal plane, when the top surface of the rail is the uppermost portion of the rail. The layers of material may be of consistent thickness. For example, each layer of material removed may be 0.2mm thick or 0.5mm thick.

Removing the layers of material at constant thickness may allow the depth of a crack to be determined. The depth of a crack may be determined by analysing the images to the point where the crack is no longer visible, and counting the layers of material that have been removed to get to that point. Removing layers of material of constant thickness allows the depth to be calculated by simple addition. The analysis of the plurality of images may indicate the direction at which the crack propagates through the rail. Comparison of co-registered images of the rail as the plurality of layers are removed will show the direction of propagation of a crack. This information may also be used to calculate the angle of propagation when combined with the depth information obtained.

The layers of material may be removed by a cutting apparatus. The cutting apparatus may be numerically controlled. The method may include the step of securing the rail to a bed of the cutting apparatus. The rail may be secured such that it cannot move during the removal of material from the top surface. The cutting apparatus may be a milling machine. The milling machine may comprise a ceramic cutting head. The layers of material may be removed from the top surface of the rail by the cutting head moving in a transverse direction. The layers of material may be removed from the top surface of the rail by the cutting head moving in a longitudinal direction.

The step of recording an image of the top surface of the rail may comprise taking a photograph of the top surface. The image may be taken by a camera in a fixed position. Taking a plurality of images from a fixed position allows the propagation of the crack through the rail to be accurately tracked as layers of material are removed. The removal of layers from the top surface of the rail may be stopped once the crack no longer propagates through the rail.

The method may include the step of adding a dye penetrant to the crack in order to increase the visual distinctiveness of the crack. The method may comprise the step of magnetic particle inspection of the crack. Such steps may assist in electronic image analysis of the crack propagation.

The method may include the step of running an eddy current sensor over the rail. The eddy current sensor may be run over the rail before any layers of material are removed from the top surface of the rail. The eddy current sensor may be run over the rail after each layer of material has been removed from the top surface. The method may include the step of calibrating the eddy current sensor in dependence on the propagation of the crack through the rail as determined by the steps of removing layers of material and image recording. The calibration step may comprise correlating the measured depth of the crack with the

response of the eddy current sensor when scanning the crack.

According to a second aspect, the invention provides a rail inspection apparatus configured for investigating the propagation of a crack through a rail comprising:

a cutting head configured to remove a plurality of layers from the top surface of a rail;

a mounting bed located below the cutting head

configured to securely fix a rail in position;

imaging apparatus configured to capture images of the top surface of a rail mounted to the mounting bed.

The imaging apparatus may be fixed in position relative to the mounting bed. The rail inspection apparatus may comprise a control means, the control means configured to control the cutting head.

The imaging apparatus may further comprise an image storage means configured to receive and store images

captured by the imaging apparatus. The control means may include the image storage means. The rail inspection apparatus may include a monitor for displaying images captured by the imaging apparatus.

The rail inspection apparatus may include an eddy current sensor.

According to a third aspect, the invention provides, a method of calibrating an eddy current sensor for inspecting rails, the method comprising the steps of:

running an eddy current sensor over a cracked rail and recording the measurement data obtained;

recording an image of the top surface of the rail, removing a layer of material from the top surface of the rail;

repeating the steps of recording an image of the top surface of the rail and removing a layer of material from the top surface of the rail a plurality of times;

analysing the plurality of images obtained to track the propagation of the crack through the rail; and

correlating the data regarding the propagation of the crack through the rail with the measurement data obtained by the eddy current sensor.

It will of course be appreciated that features

described in relation to one aspect of the present invention may be incorporated into other aspects of the present invention. For example, the method of the invention may incorporate any of the features described with reference to the apparatus of the invention and vice versa.

Description of the Drawings Embodiments of the present invention will now be described by way of example only with reference to the accompanying schematic drawings of which: Figure 1 shows a schematic view of a rail inspection

apparatus according to a first embodiment of the invention and a rail under inspection;

Figure 2 shows a schematic view of a rail inspection

apparatus as shown in figure 1 with the cutting head substituted with an eddy current sensor;

Figure 3 shows an example photograph of a top surface of a rail;

Figure 4 shows a photograph of the rail shown in figure 3 with 1.6mm of the top surface removed;

Figures 5A to 5S show a series of photographs of the top surface of a rail, with a 0.2mm layer of material removed from the top surface of the rail in each sequential figure; and Figure 6 shows an example top surface of a rail along

with the readings obtained by an eddy current sensor that has been run along the top surface of that rail.

Detailed Description

Figure 1 shows a schematic view of a rail inspection apparatus 10 according to a first embodiment of the

invention and a rail 12 under inspection. The rail 12 may include one or more cracks as will be described and shown in greater detail in the following figures. The rail 12 comprises a head 14, a web 16 and a foot 18. The head 14 comprises a top surface 20 and the foot 18 of the rail 12 is mounted to a fixing bed 22 of the rail inspection apparatus 10. The apparatus 10 further comprises a numerically controlled ceramic cutting head 24 and a camera 26 mounted in a fixed position above the rail 12. The cutting head 24 and camera 26 are controlled by a control unit 28, which also includes a memory to store images taken by the camera 26.

The cutting head 24 is arranged to move across the top surface 20 of the rail 12 in a traverse direction in comparison to the longitudinal axis of the rail 12, the traverse direction as indicated by the arrow A. As the cutting head 24 moves across the rail 12, it removes a thin section of material from the top surface 20. The cutting head 24 is 100mm in diameter, with the result that a 100mm long . section of material is removed from the top surface 20. The thickness of the layer of material removed may be set by the control unit 28 and may vary according to the

preferences of a user. In this particular embodiment, the cutting head 24 is arranged to remove 0.2mm of the top surface 20 per cutting pass. Prior to the first cutting pass, and following each subsequent cutting pass, the camera 26 is arranged to take a photograph of the top surface 20 of the rail 12. The ceramic cutting head 24 allows a very accurate and clean cut to be made across the top surface 20 of the rail as will be seen in the following figures.

The process of investigating the propagation of cracks in the rail is as follows. The camera 26 takes a photograph of the top surface of the rail. This photograph will capture the position of the cracks as they appear prior to the rail being cut. The cutting head 24 then removes 0.2mm of the top surface of the rail 12 with a first cutting pass. The camera 26 then takes another photograph of the top surface 20 of the rail, in this case showing the position of the cracks 0.2mm further into the rail.' This process ^' of cutting passes and subsequent photographing of the top surface is repeated until all of the cracks in the rail 12 have disappeared. The photographs may be used to calculate the depth of a crack, by examining the number of photographs taken before the crack disappears, and counting the number of 0.2mm sections of material have been removed. The photographs may also be used to track the direction of propagation of a crack within the rail by plotting the change of position of the cracks as the various layers of material are removed.

The apparatus as shown in figure 1 may be modified by adding an eddy current sensor as shown in figure 2. The eddy current sensor may be mounted to a standard tool receiving head on the milling machine, replacing the cutting head 24. Such an arrangement will make it easy to add or remove the eddy current sensor to the apparatus as required. Prior to the first cutting pass, the eddy current sensor may be run over the rail to detect whatever cracks are present. The rail may include copper boundary markers spaced to provide boundaries for the eddy current sensor readings. The boundaries may correspond to the part of the rail where the material will be removed by the cutting head. Once the rail has been scanned by the eddy current sensor, the crack investigation process may proceed as described above. The data provided by the crack investigation, depth and

direction of crack propagation, may be correlated with the readings obtained by the eddy current sensor. This may provide improved classification of cracks in rails when an eddy current sensor is used in rails in situ, for example, an eddy current sensor as disclosed in UK patent publication number GB 2469009.

In an alternative embodiment the eddy current sensor may be run over the top surface of the rail after every cutting pass. The plurality of eddy current readings may be correlated with the data provided by the crack investigation as previously described.

Figure 3 shows an example photographic image that may be taken by the camera 26. The top surface of the rail shows a plurality of cracks.40. The copper strips 42 act as location references for the eddy current scan as previously described. Figure 4 shows the same rail with 1.6mm of the top surface removed and the propagation of the cracks 40 through the rail.

Figures 5A to 5S show a series of photographs taken during a crack investigation process as described, with each photograph taken at 0.2mm steps. It can be seen that the curved top surface of the rail results in a small section being removed initially, with the width of the section increasing as further cutting passes are made. The depth of each crack is calculated from the cutting pass which first cuts the crack to when the crack disappears. This will prevent any errors being created by the initial curve of the top surface.

Figure 6 shows a photographic image of the top surface of a rail, alongside the readings obtained by an eddy current sensor run along the top surface of the rail. The eddy current sensor has 9 channels, each channel connected to an individual sensor, where the individual sensors are spaced transversely across the rail during the scanning processes. As can be seen, the response of. the eddy current sensor is affected by the presence of cracks in the top surface of the rail, and the investigation process detailed herein allows the information obtained about the propagation of the cracks to be correlated with the response of the eddy current sensor when run over said cracks.

Whilst the present invention has been described and illustrated with reference to particular embodiments, it will be appreciated by those of ordinary skill in the art that the invention lends itself to many different variations not specifically illustrated herein. By way of example only, certain possible variations will now be described. The rail inspection method and apparatus may be used to investigate other potential rail defects, such as squats and transverse defects.

Where in the foregoing description, integers or elements are mentioned which have known, obvious or

foreseeable equivalents, then such equivalents are herein incorporated as if individually set forth. Reference should be made to the claims for determining the true scope of the present invention, which should be construed so as to encompass any such equivalents. It will also be appreciated by the reader that integers or features of the invention that are described as preferable, advantageous, convenient or the like are optional and do not limit the scope of the independent claims. Moreover, it is to be understood that such optional integers or features, whilst of possible benefit in some embodiments of the invention, may not be desirable, and may therefore be absent, in other

embodiments .