Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
A SYSTEM AND A METHOD FOR THE DIAGNOSIS OF ABNORMAL BENDING IN RAILWAY RAILS, IN PARTICULAR AT CONNECTIONS BETWEEN RAILWAY RAILS
Document Type and Number:
WIPO Patent Application WO/2023/007273
Kind Code:
A1
Abstract:
A system (1) is described for the diagnosis of the operating conditions of joints (J) between rails (R1, R2), comprising: a frame (2) configured for installation on a railway vehicle at a wheel (W) of the railway vehicle, a first distance sensor (4) and a second distance sensor (6) installed on said frame (2) and arranged in such a way as to be on opposite sides of a wheel (W) of the railway vehicle, each of said first distance sensor (4) and second distance sensor (6) being configured to detect, respectively, a first distance and a second distance between said frame (2) and a top of a rail (R1, R2), a third distance sensor (8) installed on said frame (2) and configured to detect a third distance of said frame (2) with respect to a journal box (10) associated to the wheel (W) of the railway vehicle.

Inventors:
SANTANERA CESARE (IT)
Application Number:
PCT/IB2022/055949
Publication Date:
February 02, 2023
Filing Date:
June 27, 2022
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
DMA S R L (IT)
International Classes:
B61K9/08; B61L15/00; B61L23/04; E01B35/06
Foreign References:
JPH06235609A1994-08-23
US20080228436A12008-09-18
Attorney, Agent or Firm:
DE BONIS, Paolo (IT)
Download PDF:
Claims:
CLAIMS

1. A system (1) for the diagnosis of bending of railway rails (Rl, R2, J), comprising:

- a frame (2) configured for installation on a railway vehicle at a wheel (W) of the railway vehicle, a first distance sensor (4) and a second distance sensor (6) installed on said frame (2) and arranged so as to be on opposite sides of a wheel (W) of the railway vehicle, each of said first distance sensor (4) and second distance sensor (6) being configured to detect, respectively, a first distance and a second distance between said frame (2) and a top of a rail (Rl, R2)

- a third distance sensor (8) installed on said frame (2) and configured to detect a third distance of said frame (2) with respect to a journal box (10) associated to the wheel (W) of the railway vehicle.

2 . The system (1) according to Claim 1, comprising at least one further sensor (12, 14, 16) configured to determine, on the basis of said first distance (a), second distance (b) and third distance (f), a spatial trajectory of said frame (2) referred to a top of said rail (Rl, R2).

3. The system (1) according to Claim 2, wherein said at least one further sensor comprises at least one inertial sensor (12, 14, 16) installed on said frame

(2).

4 . The system (1) according to Claim 2, wherein said at least one inertial sensor comprises a first accelerometer (14) installed on said frame (2) and a second accelerometer (16), each of said first accelerometer (14) and second accelerometer (16) being accelerometers configured to detect a vertical acceleration.

5 . The system (1) according to Claim 4, wherein said first accelerometer (14) is installed on said frame (2) in correspondence of said first distance sensor (4), and wherein said second accelerometer (16) is installed on said frame (2) in correspondence of said second distance sensor (6).

6. The system (1) according to Claim 2 or Claim 3, wherein said at least one inertial sensor comprises an inertial platform (8) installed on said frame (2).

7 . The system (1) according to any one of the preceding claims, further comprising a locationing unit (18) configured for the spatial referencing of the distance detected by said at least one distance sensor (4, 6, 8), preferably for the time and/or spatial referencing of the distance detected by said at least one distance sensor (4, 6, 8).

8. A method for diagnosing bending of railway rails (Rl, R2) by means of a system (1) according to any one of the preceding claims on board a railway vehicle, the method comprising:

- running the railway vehicle along said track in order to go through an inspection area (J), detecting, during the running of the railway vehicle, said first distance (a) at a first point (PI), said second distance (b) at a second point (P2), and said third distance (f) at a third point (P3), said first, second and third points (PI, P2, P3) being located on the top of a rail, determining a figure having vertices at said first, second and third points (PI, P2, P3) determine the values of a first angle (oil) defined between a side of said figure joining the first point and the second point (P1-P2) and a side of the figure joining the first point and the third point (PI PS), of a second angle (cx2) defined between a side of said figure joining the first point and the second point (P1-P2) and a side of the figure joining the second point and the third point (P2-P3),

- determining a fourth distance value (m) between said third point (P3) and the side of said figure joining the first point and the second point (P1-P2),

- comparing the values of said first angle (oil), second angle (cx2) and fourth distance value (m) with respective threshold values (m_0, al_0, a2_0).

9. The method of Claim 8, further comprising correcting the values of said first distance (a, Ga), said second distance (b, Gb) and said third distance (f, Gf) on the basis of the displacements of the frame (2) with respect to the rail determined by means of said inertial sensor (12, 14, 16), and determining:

- a first difference between a corrected value of the first distance and a corrected value of the second distance (Ga-Gb),

- a second difference between a corrected value of the first distance and a corrected value of the third distance (Ga-Gf)

- a third difference between a corrected value of the second distance and a corrected value of the third distance (Gb-Gf).

10 . The method of Claim 8, further comprising comparing said first difference (Ga-Gb), second difference (Ga-Gf) and third difference (Gb-Gf) with respective threshold values.

Description:
"A system and a method for the diagnosis of abnormal bending in railway rails, in particular at connections between railway rails"

★★★★

TEXT OF THE DESCRIPTION

Field of the Invention

The present invention refers to the diagnosis of railway facilities. More specifically, the invention has been developed with reference to the diagnostics on railway rails.

Prior art and General Technical Problem

The railway lines have countless connection joints between adjoining rails, which ensure the geometric and structural continuity of the railway line. The joints between rails are traditionally a point of weakness of the track, both if they are implemented by welding or by means of joint plates bolted to the rail web.

The progressive damage and the failure of joints typically take place due to the rail bending, caused by the weight applied by the wheels of railway vehicles. If the rail is not well supported by the ballast in that point, the vertical load generates a local bending, and therefore a fatigue that in the long run will cause the joint plates or the welding to break (especially if there are manufacturing flaws in the components). The same phenomenon may take place along the continuous rail (i.e., in a section other than the joints) in case of a local lack of support, and the risk of failure increases if, for any other reason, cracks are already present.

Without questioning the usefulness of dedicated diagnosis vehicles, it is obvious that the diagnosis carried out by such vehicles has the great disadvantage of being executed non-continually, and generally speaking with a frequency which is insufficient to detect very rapid degradation phenomena: the passage of a railway vehicle dedicated to the diagnosis occupies a slot of the line track subjected to the inspection, and this has a negative impact on the general railway circulation; moreover, such a vehicle may have features (e.g. weight per axle) which are not comparable to those of the railway vehicles normally running on the line. In the latter case, the diagnosis may be biased by the diagnostic vehicle itself:in the case of a weight per axle which is sensibly lower than in the normal railway vehicles, the phenomena of rail bending at the passage of the railway vehicle may be dangerously underestimated.

Object of the Invention at the object of the invention is to solve the technical problems outlined in the foregoing. Specifically, the object of the invention is diagnosing, in advance and continually, the existence of degradation conditions in the track at the joints, or in any other place where a local fatigue stress (due to bending on the vertical plane) may occur, so as to enable taking action before dangerous situations arise.

Summary of the Invention

The object of the invention is achieved by means of a system and a method having the features set forth in the claims that follow, which form an integral part of the technical disclosure provided herein in relation to the invention.

Brief Description of the Figures

The invention will now be described with reference to the annexed Figures, which are provided by way of non-limiting example only, and wherein:

Figure 1 schematically shows a diagnostic system according to various embodiments of the invention, and - Figures 2, 3A, 3B show aspects of a diagnostic method according to the invention.

The Cartesian reference system shown in the Figures identifies the longitudinal direction - axis X, the transversal direction - axis Y and the vertical direction - axis Z.

Detailed Description

Reference 1 in Figure 1 generally denotes a system for the diagnosis of abnormal bending in railway rails. Figure 1 schematically shows the system 1 positioned on a pair of rails R1 and R2, which are connected at a joint J which in the present case consists of bolted plates. Rails Rl, R2 form a part of a railway track, mounted on sleepers C and supported by railway ballast BL.

It will be observed that, although the Figures show a joint between rails which is implemented by means of bolted plates W, the same method applies to welded joints and even to a continuous rail, if a discontinuity is encountered in the support provided by ballast BL to sleepers C and therefore to the rails. Therefore, joint J must be understood as representative of a possible inspection area of the rail.

The description in the following will refer for brevity only to the case of the diagnosis of bending of joints J, being it understood that it may also apply to continuous rails without joints. System 1 is generally configured for the installation on a railway vehicle, whatever the type thereof may be: freight wagons, passenger cars, locomotives, provided that they have a sufficiently loaded axle. The diagnostic system 1 comprises a substantially rigid frame 2, which acts as a reference for all measurements made by system 1. Frame 2 may be a frame of the bogie of the railway vehicle whereon the system 1 is installed, or else it may be an auxiliary frame which may be installed onto a bogie frame or below the body of the railway vehicle itself. Frame 2 may also correspond to the very frame of the railway vehicle, especially in the case of two- axle vehicles.

In preferred embodiments, frame 2 extends from one side to the other side of the railway vehicle, i.e. it substantially covers the whole gauge between the rails. In other embodiments, frame 2 may extend transversally along a part of the gauge, and has a twin configuration (two frames 2, each being associated to each wheel of the wheelset) . As will be more apparent in the following, the advantage of having a frame 2 extending to one side to the other of the gauge consists in using only one inertial platform at the middle of frame 2, instead of two inertial platforms respectively above each rail Rl, R2.

In various embodiments, the system 1 comprises a set of sensors, which may vary according to the different embodiments of the diagnostic method according to the invention.

In an embodiment with full equipment, system 1 comprises: a first distance sensor 4, preferably a contactless sensor (in any case it is also possible to use a contact sensor: it may be a LVDT transducer or other displacement sensor performing the same functions), configured to detect a distance a from the top (the upper edge) of rail Rl and the frame 2, specifically with respect to the position which the sensor occupies on the frame 2. Distance a is measured along a vertical direction Z; in the diagnostic method according to the invention, it is possible to take into account either the absolute value of the distance or, preferably, the variation of distance a with respect to a value which may be detected when the railway vehicle is standing still on a track section which is continuous and in very good conditions, i.e., which bends under the vehicle load by a measure corresponding to what the design envisages. a second distance sensor 6, preferably a contactless sensor (in any case it is possible to use contact sensors), configured to detect a distance b from the top (the upper edge) of rail R2 and the frame 2, specifically with respect to the position which the sensor occupies on the frame 2. Distance b is measured along a vertical direction Z; in the diagnostic method according to the invention, it is possible to take into account either the absolute value of the distance or, preferably, the variation of distance b with respect to a value which may be detected when the railway vehicle is standing still on a track section which is continuous and in very good conditions, i.e., which bends under the vehicle load by a measure corresponding to what the design envisages.

It must moreover be remarked that distance sensor 6 is arranged on an opposite side of joint J with respect to sensor 4 when the wheel W exerting its load on the rail is on joint J or on the bending weak point of the rail: in other words, sensors 4 and 6 (which, unlike wheel B, are obviously unable to exert forces) are positioned on the frame 2 so as to be located astride the joint J or the weak point, when the wheel exerts a load onto it and makes it bend. a third distance sensor 8, preferably a contactless sensor (in any case it is possible to use a contact sensor: it may be a LVDT transducer or other displacement sensor performing the same functions), configured to detect a distance f from the frame 2 to a journal box 10, which engages a wheel W of the railway vehicle whereon the system 1 is installed. Because, obviously, the journal box 10 and the wheel W keep their relative position in all conditions (if not, the wheelset would break), the detection of distance f from the journal box corresponds, with the subtraction of a fixed value corresponding to the dimension of the journal box with reference to the centre of wheel W, to the detection of the position of the centre of wheel W with respect to the frame 2. What is detected, as will be apparent from the description in the following, is the variation of the relative position between wheel W and frame 2 which takes place between the following conditions: i) the railway vehicle is standing still on a continuous track section which is in very good conditions (i.e., which undergoes bending phenomena within the design specifications - said position is represented by an ideal profile R_ID of the rail sequence Rl, R2), and ii) the vehicle wheel W exerts a load onto joint J or a rail Rl, R2 which is not supported by ballast.

It is optionally possible to provide the system 1 with an inertial platform 12 installed on the frame 2, if one single frame 2 is present, or generally speaking with an inertial platform 12 for each frame element composing frame 2. The inertial platform 8 provides the trajectory of the frame 2 whereon it is installed, and enables defining a spatial reference line 2R which corresponds to an ideal reference trajectory of frame 2 in space. It must be remarked that the inertial platform 8 may be installed nearly anywhere on frame 2: the shown position must be understood as merely exemplary, because algorithms are known to manage the different relative positions of the sensors and of the inertial platform.

Optionally, again, it is possible to provide the system 1 with a first vertical accelerometer 14 (which is therefore adapted to detect accelerations along the vertical direction Z) and a second vertical accelerometer 16 (which is therefore adapted to detect accelerations along the vertical direction Z) respectively installed at the positions of sensor 4 and of sensor 6. The accelerometers 14 and 16, in some embodiments, may replace the inertial platform 12, at a slightly lower cost (but of course without the additional performances which the inertial platform 8 may offer).

Further components of system 1 comprise:

- a locationing unit 18, preferably including an odometer having a millimeter resolution or higher. The locationing unit 18 enables identifying all the measurements made by system 1, i.e., associating the measurement to the instant (in time and/or in space) of acquisition detected by the odometer. In this regard, various methods are known and widespread for associating the value detected by the odometer to geographical (e.g., GPS) coordinates to the kilometre stakes present in the network; the odometer (and, generally speaking, the locationing unit 18) is moreover configured to be used as a trigger for the equally spaced acquisition of the data of sensors 4, 6 and 8, and for the spatial referencing of the data from inertial platform 8 and from accelerometers 14, 16, even though the latter are acquired at a constant time (and not spatial) frequency.

Optionally, again, it is possible to envisage an automatic system 20 for identifying joints J. The system may be implemented by means of any known technique, e.g., through the automatic recognition of images.

In any case it must be observed that, although it is important to locate the joint, preferred embodiments of the diagnostic method according to the invention envisage keeping system 1 constantly active, as if joints J were present everywhere. The presence of the joint or of the weak point of the rail is detected when the conditions described in the following are met. Such conditions might appear also because of a failure of the rail, or due to the presence of a bending weak point which will generate a fatigue failure, and which of course may appear at any position: therefore, irrespective of the implementation and/or of the presence of the automatic system 20, system 1 is preferably always active, and not only active where a joint is known to be present.

At any rate, generally speaking, it may be remarked that the automatic system 20 for identifying joint J is useful for detecting defective joints, broken rails and weak points,

- monitoring all the joints, including those in good conditions, and generate a report, if this is useful; obviously, a perfect joint is equivalent to a continuous rail, and therefore it cannot be identified through its defects.

By way of non-limiting exemple only, the following Table identifies some preferred embodiments of system 1, which differ from one another in their configuration. The first row of the following table indicates the reference number corresponding to the sensors or to the components described in the foregoing, and the following rows identify respective embodiments and, in the presence of character "X", the presence of the sensor or of the component.

Sensors 4, 6, 8 have a sub-millimetre uncertainty of measurement (preferably 0,1 mm) . The required uncertainty of measurement also depends on the quality of the rails being inspected and on the desired margin of safety.

The quality of the inertial platform 12 and of the accelerometers 14, 16, especially as regards the signal-to-noise ratio and drifts and as regards temperature stability, depends on the desired sensitivity for carrying out preventive diagnostics and on the minimum speed at which it is desired to operate. Generally speaking, sensors are preferred which are adapted to perform integrations of at least 10 seconds without producing an error higher than a fraction of a mm.

The signals coming from sensors 4, 6, 8 are preferably sampled at least every 125 mm, even though it is possible to operate with a slightly lower sampling rate (i.e., at longer intervals). At any rate, the measurement is more reliable if sampling takes place every 25 mm or less, in order to detect with accuracy the instant when the load is centered on the weakest point of the rail, and therefore the maximum difference is generated between the data acquired by the sensors 4, 6 (without load) and the sensor 8, which detects the effects of the load.

The signals from the inertial platform or from the accelerometers are sampled in time, for known reasons, and are then sampled again in space within the calculation sequence. A more frequent sampling also enables filtering the noise out of the signal, even though it is preferable to employ sensors having a low level of noise.

The operation of system 1 according to the first embodiment of the invention, as shown in the previous Table, will now be described. The second and the third embodiments are configured as substantially more economical versions of the first embodiment, and as such they do not enable carrying out all the determinations and the deductions of the first embodiment.

Referring to Figure 3A, the distance sensors 4, 6, 8 are set to zero - i.e., their zero reference is established, which may be equal to a reference distance value - once for every operation cycle of system 1, and setting to zero is repeated during maintenance. The setting to zero takes place on a straight track section which is in good conditions and of good quality (as regards construction and geometry).

The setting to zero consists in determining, for each sensor, an additional value 54, 56, 58 (offset - in sum or in subtraction) which makes the readings a, b, f of sensors 4, 6, 8 identical to each other; such readings are always referred to the upper edge of rails R1 and R2 for each of the sensors 4, 6, 8. The term "upper edge" indicates the highest point of the top of the rail: in the following, for brevity, the term "edge" will often be used.

A possibility consists in determining the offsets 54, 56, 58 in such a way as to have all readings a, b, f equal to zero; however, another possibility consists - with reference to Figure 3A - in determining the offsets 54, 56, 58 with respect to reference line 2R, so that the algebraic sums thereof and, respectively, the readings a, b, f are identical. Therefore, a corresponding number of offset values are obtained, which will always be summed algebraically to the value measured by each sensor in order to provide a correct output; with such assumptions, in the following part of the description, the distance data and the readings from sensors 4, 6, 8 mentioned will always be readings corrected by the offset determined in this fashion. The use of the offset values 54, 56, 58, moreover, enables using the readings a, b, f as references for the diagnostic activity on the rails provided by the invention: the calibration of the offsets enables defining, in practice, the condition of absence of failure (a sort of zero reference for the reading, irrespective of whether the offset calibration is implemented by setting to zero the resulting signal): if the rail or the joint J does not undergo significant bending phenomena under the load of wheel W, the readings of system 1 will always be such as not to generate significant differences from the values which locate the aligned points Pi (distance a), P2 (distance b), P3 (distance f). Each point Pi (in the present case i = 1, 2, 3) corresponds to a point on the edge (on the top) of rail Rl, R2 at which the respective distance a, b, f is detected. On the contrary, if a significant bending phenomenon is generated, a deviation from the reference condition is observed which may then be rapidly and easily diagnosed by system 1: in this respect, see the following description for further details.

All the inertial sensors (platform 12 and accelerometers 14, 16) are calibrated by means of known techniques. The readings a, b, f provide the distances of the frame from the rail(s) Rl, R2: the inertial platform 12 (as well as the accelerometers 14, 16, which integrate the signal thereof in time) is adapted to detect the trajectory of frame 2, but does not know the position thereof with respect to the rails: by referring - by means of the offsets 54, 56, 58 - the readings a, b, f to the trajectory of the frame 2, which is detected by way of the inertial platform 12, it becomes immediately possible to detect the shape on the vertical plane of the rail(s) Rl, R2.

The element common to all the embodiments of system 1 and of the diagnostic method according to the invention involves the continuous equally spaced detection of the vertical distances a, b, f referred to the upper edge of the rails Rl and R2 for each of the sensors 4, 6, 8.

Referring to Figures 1, 2 and 3B, while the railway vehicle is running with speed V along the rails Rl, R2, crossing joint J, a series of distances hi is obtained (i.e., as previously stated, the distances a, b, f corrected according to the calculated offset) for each of the sensors 4, 6, 8. At the same time the acquisition takes place of the data of the inertial platform 12, which are processed in real time according to known algorithms, in such a way as to obtain the positions and the attitude of frame 2 in the 3D space at the same points where the values a, b, f and the related distances hi are measured. In Figure 3B, the distances a, b, f are associated to an apex (i.e., a', b', f ) in order to highlight the fact that they are detected on a rail which sags by bending. The algorithms for correlating the data of the inertial platform with the data acquired by sensors 4, 6, 8 are per se known and will not be described in detail.

From the processing of the above data, four continuous functions are obtained (the variable X being the distance covered along the track comprising the rails R1, R2):

- Ga = f(X) Vertical geometry (plane XZ) of the rail, calculated by both inertial platform 12 and sensor 4 (distance a, with correction; the role of the inertial platform has been described in the foregoing and comprises creating a longitudinal reference, represented by line 2R)

- Gb = f(X) Vertical geometry (plane XZ) of the rail, calculated by both inertial platform 12 and sensor 6 (distance b, with correction; the role of the inertial platform has been described in the foregoing and comprises creating a longitudinal reference, represented by line 2R)

- Gf = f(X) Vertical geometry (plane XZ) of the rail, calculated by both inertial platform 12 and sensor 8 (distance f, with correction; the role of the inertial platform has been described in the foregoing and comprises creating a longitudinal reference, represented by line 2R)

- F = f(X) Deflection calculated from the three values of corrected distances a, b, f. Substantially, a straight line is traced passing through points Pi on the upper edge of the rail, corresponding to the readings of the distances a and b (Pi and P 2 in the

Figure; i = 1, 2). Then, it is checked how much the point Pi on the upper edge of the rail, defined by distance f (P 3 in the Figure; i = 3) is displaced from this straight line. This displacement will be denoted as "deflection" F in the following.

The (vertical) geometry of the rail, which derives from the above functions, is preferably calculated on rail lengths (or wave lengths) which are much shorter than the traditional measures, specifically as little as 50 cm, or even less.

This means that the present approach is different from the traditional measurements, both as regards the much more frequent spatial sampling, and as regards the filters which extract the measurements on short wave. It must be observed, moreover, that by using known algorithms and thanks to the fact that frame 2 is rigid, by means of one single inertial platform 12 it is possible to calculate the geometry of the track in three different points of the carrier vehicle (of course, if frame 2 comprises one single element).

Then, the points Pi are determined whereat the value of function F exceeds a threshold value, which generally depends on the type of track and of traffic (which however is at most of the order of magnitude of the millimetres, approximately 1 to 3 mm).

In the area of joint J (and the same applies to welded joints or continuous rail lengths), the functions Ga, Gb, Gf are substantially identical in normal conditions, i.e. in bending conditions of the joint J or of the rail which correspond to the design requirements .

Therefore, there are defined at least a first (pre-alert) threshold and a second (alert) threshold for the maximum value of the module of differences (Gf-Ga), (Gf-Gb), (Gb-Ga), calculated on a length of approximately two metres or more across joint J (or, generally speaking, along the rail section which is being inspected).

The differences (Gf-Ga), (Gf-Gb) represent the amount of sagging of the rail under the load applied by wheel W. An excessive sagging is a reason for alarm, especially when the points of installation on frame 2 of sensors 4 and 6 for measuring distances a and b are close to the arrangement position of sensor 8 for measuring distance f.

The difference (Gb-Ga) should generally be close to zero, excluding the measurement errors and a very small hysteresis which also depends on the speed of the railway vehicle. If the difference (Gb-Ga) exceeds a respective threshold value, this indicates a hysteretic (anelastic) behaviour, which beyond given limits is a reason for alarm in itself.

Subsequently, the calculation of distances a, b, f is performed at the moment when the maximum value of function F is detected, i.e., when the bending at joint J or at the point of the rail being inspected is maximum.

Referring to Figure 3, by means of the system 1 and of the diagnostic method according to the invention it is possible to measure the values m, oil, 2 depicted in Figure 3B, i.e.: m: maximum deflection of joint J under load (that is the maximum of function F) oil: displacement angle of rail R1 (or of the rail section) with respect to the reference profile R_ID, which in the present case is the line joining the points on the upper edge of the railway identified by a and b. Therefore, this is one of the acute angles at the base of the triangle defined by points Pi (a), P2 (b), P 3 (f). 2: displacement angle of rail R2 with respect to the reference profile R_ID, which in the present case is the line joining the points Pi, P2 on the upper edge of the rail identified at the distances a and b. Therefore, this is another acute angle at the base of the triangle defined by points Pi (a), P2 (b), P 3 (f).

The values are calculated according to known trigonometric functions, which are not stated herein for brevity.

The existence is therefore determined of a potentially critical condition in joint J (or generally at the point being inspected), if one or more of the following conditions are met: i) m > mO ii) oil > al_0 iii) 2 > a2_0 iv) oil + 2 > al2_0 wherein mO, al_0, a2_0, al2_0 are respectively threshold values for the parameters m, oil, cx2, oil + OL2.

Substantially, thanks to system 1 and to the diagnostic method according to the invention, an excessive bending of the rails under the load is to be considered defective, especially if it appears along a short inspection length, and therefore with a high curvature and as a consequence with a remarkable fatigue of the material.

As far as the second and third preferred embodiments shown in the Table above, the following considerations apply:

Second embodiment: deflection only

The functions Ga, Gb, Gf cannot be calculated due to the absence of inertial sensors (inertial platform 12 and/or accelerometers 14, 16). It is however possible, thanks to the distance sensors 4, 6, 8 and the calibration thereof, to determine the values m, oil e oi2, which are in themselves an important diagnostic tool.

Third embodiment: simplified inertial

In this embodiment of system 1 it is not possible to operate in three dimensions in space, due to the absence of the inertial platform; therefore, the system 1 can only operate on the basis of coordinates in the plane XZ by means of the vertical accelerometers 14, 16. By means of a double integration in time of the signal of accelerometer 14, 16, the movement in plane XZ of frame 2 is calculated, so as to obtain, albeit with a slightly higher error than in the case of the first embodiment, all the functions Ga, Gb, Gf, F which may be calculated with the system according to the first embodiment. The calculation of parameters m, oil e 2.

Obviously, in this context it is of paramount importance that the sensors 4 and 6 are arranged at the correct distance from wheel W, and that all the sensors are very accurate: because the angles are measured by means of a "short" segment, a small error of a, f, b causes a rotational error of the segment and therefore of the angle measurement.

To sum up, all the embodiments according to the invention define a method for the diagnosis of the bending of joints J between rails Rl, R2 or weak points of a continuous rail by means of the system 1 installed on board a railway vehicle, wherein the method comprises:

- running the railway vehicle (in the Figures, V indicates the longitudinal speed of the vehicle) along a track - which comprises a pair of rails, which are either continuous or interrupted by joints J, and each of which may be inspected by means of system 1 - in order to go through an inspection area (J, or an area of continuous rail whereat a ballast sagging is present), detecting, during the running of the railway vehicle, the first distance a at a first point PI, the second distance b at a second point P2, and the third distance f at a third point P3, said first, second and third points PI, P2, P3 being located on the top of the rail, determining a figure having vertices at said first, second and third points PI, P2, P3, determining the values of a first angle oil defined between a side of said figure joining the first point and the second point (P1-P2) and a side of the figure joining the first point and the third point (PI PS), of a second angle 2 defined between a side of said figure joining the first point and the second point (P1-P2) and a side of the figure joining the second point and the third point (P2-P3),

- determining a fourth distance value m between the third point P3 and the side of the figure joining the first point and the second point (P1-P2),

- comparing the values of said first angle oil, second angle 2 and fourth distance value m with respective threshold values m_0, al_0, a2_0.

Out of the three vertical distance measurements performed by the sensors 4, 6, 8, one (f) is executed under the load trasmitted by the rail wheel W, while the others are performed at both sides (upstream and downstream in the longitudinal direction X) at a distance of approximately 50 and 100 cm with respect to the axis of wheel W, i.e., with respect to the point of acquisition of distance f. The measurements at the sides are performed without an appreciable vertical load, i.e., ideally contactless, by means of any available technology.

The skilled person will appreciate how, by means of the system according to the invention, it is possible to solve the problems of the prior art by diagnosing in advance, and continually, the existence of degradation conditions in the joints J between the rails or in continuous rails which are poorly supported by ballast BL, so as to take action before dangerous situations arise. It is not even necessary to resort to diagnostic vehicles, because the system 1 may be installed on board railway vehicles which normally operate for passenger or freight services along railway lines, therefore obtaining the further advantage of performing the measurements under real load conditions of the rails, which may not always be reproduced by means of a dedicated railway vehicle. The wheel W itself of the railway vehicle operates as a measuring element, because the wheel directly imparts the vertical load to rails Rl, R2 at the joint J or at the weak point of the rail. Once the weak point has been located, whatever the nature thereof may be, the calculations are performed and the inferences are drawn as described above. The proposed method offers the further advantage of being adapted to be installed on vehicles where no person performing a diagnostic activity is present. Generally speaking, the only requirement for the method according to the inventions concerns the weight per axle of the vehicle carrying system 1, which must be close to that of the commercially operating vehicles, so as to exert on the rails the same load they would be subjected to during the normal running of such commercial vehicles.

Of course, the implementation details and the embodiments may amply vary with respect to what has been described and illustrated, without departing from the scope of the invention as defined by the annexed claims.