RU2712975C1 - Method of ultrasonic detection of longitudinal cracks in a rail head - Google Patents

Abstract

FIELD: defectoscopy.

SUBSTANCE: use for ultrasonic detection of longitudinal cracks in rail head. Summary of invention consists in the fact that first electroacoustic transducer is mounted on rail rolling surface, oriented by deflection of rail head to lower fillet of head on side of working face with possibility of receiving reflected signals from it, moving converter along rail at a certain rate, periodically emitting ultrasonic vibrations, receiving echo signals reflected from lower head fillet on side of working face of rail, analyzing amplitude of received signals, at that echo signals reflected from lower and upper planes of longitudinal crack are received, first electroacoustic transducer is mounted mirror-wise to the first, but directed to the lower fillet on the side of the non-working face of the rail head and functions similarly to the first one, together with the first second transducer is moved along the rail, evaluation of depth of occurrence and degree of longitudinal crack development is carried out on the basis of analysis of amplitudes and time position of echo signals from a crack and levels of amplitudes of echo signals from bottom fillets of a rail head in series fed as transducers move.

EFFECT: rails reliability and efficiency, while simplifying implementation of the method.

1 cl, 6 dwg, 1 tbl

Description

The invention relates to the field of ultrasonic (US) non-destructive testing of railway rails and can be used to detect and evaluate longitudinal cracks in the rail head.

61% of all defects found in recent years in rails on the road network of Russian Railways are horizontal longitudinal cracks. Arising on the working fillet of the rail head, extended horizontal contact fatigue cracks propagate deep into the rail head as it develops. If we take into account that about 70% of defects are formed in the rail head, then among them 80% are longitudinal cracks and only 20% are transverse cracks in the head. These figures are confirmed by both our research and the VNIIZhT publications.

The current regulatory documentation [1] on the degree of danger considers two groups of extended cracks in the rail head: with a depth of location from 2.0 to 8.0 mm (defects in codes 10, 11 and 12); and with a depth of 8.0 mm or more from the ski surface (code 30.1-2). The latter, regardless of the length and zone of the defect along the rail, are classified as dangerous (acute defective rail - ODR) and must be replaced without delay. And subsurface longitudinal cracks (up to 8.0 mm) are classified as dangerous only when their length in the zone of the longitudinal axis of the rail exceeds 70 mm. Moreover, the wider (across the rail) the crack propagates, the more dangerous the defect.

Thus, the task of detecting and measuring the parameters of longitudinal horizontal cracks in the rail head is relevant. In this case, it is very important to be able to separate them according to their depth: up to and more than 8.0 mm.

Analysis of defectograms with the indicated defects shows that the existing sounding schemes implemented in modern flaw detectors do not provide the formation of clear and unambiguous signals from the desired defects. As a result, their timely detection is difficult and requires considerable experience from decryptors. Of 11 -12 ultrasonic (ultrasound) channels of flaw detection systems, only one channel with an electro-acoustic transducer (EAP) with an input angle of 0 ° (direct transducer) using longitudinal ultrasound waves is designed to detect the cracks under consideration.

However, a direct transducer can only detect cracks “entering” into the projection of the neck (i.e., sufficiently developed). In addition, the depth of occurrence of cracks (5 -15 mm) is usually located in the "dead" zone of the transducer, which prevents a clear echo from the first reflection. Indeed, when using a direct EAP, longitudinal waves are emitted at a speed of 5.9 mm / μs, which leads to the echo from the crack entering the so-called “dead” zone (the echo enters the zone of operation of the probe pulse) and impossibility, in most cases, assessing the depth of a longitudinal crack. For example, when a transverse crack occurs at a depth of h = 12 mm from the rolling surface of the rail, the temporary position of the echo signal from the plane of the crack is t = (2 x 12) / 5.9 = 4.1 μs. At a typical frequency of ultrasonic vibrations of 2.5 MHz, used for rail flaw detection, the duration of the probe pulse is about 4.0-5.0 μs (10-12 periods at the emitted frequency). Due to transients in the protector of the EAP and in the contact layer, the actual duration is longer. As a result, the echo signal from the crack merges with the probe pulse.

Possible reflections from the edges of cracks during oblique entry of ultrasonic vibrations due to the formation of diffracted waves, as a rule, have insignificant amplitudes and are detected uncertainly on the working sensitivities of the flaw detector, against the background of noise (interference) that are unavoidable during practical control.

Basically, the detection of longitudinal horizontal delaminations is carried out only due to multiple re-reflections of ultrasonic vibrations from the crack plane and the inner surface of the rail. However, as shown by the results of forced dolomites of rails with the cracks under consideration, many of them can hardly be called "horizontal", because during the development of a crack from the working fillet deep into the head section, they acquire a certain inclination (10 ° - 20 °) relative to the horizontal. As a result, the signals reflected from the plane of the crack do not return to the radiating-receiving EAP.

The analysis shows that the current sounding schemes of modern flaw detectors and systems lack effective schemes and methods that reliably and unequivocally detect longitudinal cracks in the rail head, which make up more than half of all detected hazardous defects on the Russian Railways road network.

Most of the known ultrasonic (US) methods for detecting defects in the rail head are focused on the detection of transverse cracks (in foreign patents of “oval-shaped cracks”) [2-5], (in Russian - defects of codes 21.1-2; 20; 24 and 26.3 [1]) [6-7].

Thus, a paradoxical situation was created. All the efforts of developers and inventors in recent years, both in Russia and abroad, were mainly aimed at identifying transverse head cracks, which currently make up only 10-20% of all detected defects in the head. At the same time, the most widespread defects in the form of longitudinal horizontal cracks by known methods are uncertainly detected, which leads both to the omission of dangerous defects and to the rejection of serviceable rails.

There are known ultrasonic methods for detecting surface microcracks [8-10], originating in the contact zone of the wheels of the rolling stock with the rolling surface of the rail head (from the side of the working face) and subsequently spreading along the surface to the middle of the head and beyond. Known methods do not allow reliable detection and unambiguous determination of the degree of development of longitudinal cracks propagating deep into the rail head.

The closest technical solution adopted for the prototype is the method of ultrasonic detection of microcracks on the tread surface of the rail head according to the patent [11], which means that ultrasonic vibrations directed along the rail at certain angles are introduced from the tread surface relative to the plane of the cross section of the rail, with a line of receiving EAP located in the wheel converters, receive vibrations reflected from the lower fillet of the rail head, taking into account the signals received by the receiving converters in rulers and additional EAP, inputting and receiving ultrasonic vibrations across the head in the direction of the lower fillet of the working face of the rail head, judge the presence and degree of development of microcracks on the surface of the rail. When implementing the known method, they mainly use the shadow method of ultrasonic testing, in which the weakening of the vibrations transmitted through the rail head along an inclined path along the rail with reflections from the lower plane of the head of the ultrasound indicates the presence of obstacles in the form of microcracks in the way of their propagation. An additional input and reception of ultrasonic vibrations across the rail head is intended to control the position of the ultrasound system relative to the longitudinal axis of the rail and the quality of the acoustic contact between the rolling surface and the EAP of the emitting system.

The disadvantages of this method are:

- low reliability of the control, caused by the inability to determine the depth of longitudinal cracks, as the basis of the known method based on the shadow principle of detection of microcracks. In addition, at high control speeds (large revolutions of wheel rotation) air bubbles appear in the immersion fluid inside the wheel, creating a significant level of acoustic noise and further reducing the reliability of the control;

- low control performance caused by limited scanning speed. In the known scheme, an inclined input of ultrasonic vibrations is used, directed at a certain angle along the rail head. In this case, the travel time of ultrasonic vibrations from the emitter to the receiving transducers is more than 100 μs. In a wheel EAP, to the travel time of ultrasonic vibrations in a controlled rail, the travel time in immersion fluid in the wheel is added. In this case, the total propagation time of ultrasonic vibrations from the emitter (through the section of the rail head) to the receivers is significant, which does not allow to increase the frequency of sending ultrasonic vibrations and limits the scanning speed.

- difficulties in the implementation, requiring the use of a complex emitting-receiving system, consisting of emitting EAP slides and lines of receiving EAP placed in two wheels with an elastic shell filled with a fluid that transmits ultrasonic vibrations.

The tasks solved by the claimed method is to increase the reliability and performance of the control of the rail head while simplifying the implementation of the method.

To solve the problems, in the method of ultrasonic detection of longitudinal cracks of the rail head, which consists in the fact that the first electroacoustic transducer is installed on the rolling surface of the rail, the oriented correction of the rail head to the lower fillet of the head from the side of the working face with the possibility of receiving reflected signals from it, the transducer is moved along the rail at a certain speed, periodically emit ultrasonic vibrations, receive reflected from the lower fillet of the head from the working side the echoes of the rail face are analyzed, the amplitude of the received signals is analyzed, the echoes reflected from the lower and upper planes of the longitudinal crack are additionally received, the second electroacoustic transducer is directed to the first, directed to the lower fillet from the side of the non-working edge of the rail head, and functions similarly to the first, jointly with the first transducer, the second transducer is moved along the rail, an assessment of the depth and degree of development of a longitudinal crack is carried out on the basis of analysis a the amplitudes and temporary position of the successive arriving, as the transducers move, echo signals from the crack and the echo signal amplitude levels from the lower fillets of the rail head, in addition, the periods of emission of ultrasonic vibrations by the transducers are selected based on the required resolution to estimate the lengths of longitudinal cracks and their speed displacement.

The main idea of the proposed method is the simultaneous implementation of echo- and mirror-shadow methods of ultrasonic testing [12] to evaluate the parameters of a longitudinal crack in the rail head. In the process of moving along a rail of two inclined EAFs that are mirror-oriented transverse to the rail head, the crack is successively voiced from above and below its plane. Based on the received echo signals from the crack (echo method) and analysis of the amplitudes of the echo signals from the lower fillets of the rail head (mirror-shadow method), the parameters (depth of occurrence, crack width in plan and length along the rail length) of the desired crack are judged.

The technical result of using the proposed method is to increase the reliability and performance of the control of the rail head while simplifying the implementation of the method.

Significant differences of the proposed method in comparison with the prototype are the following actions.

1. Two EAPs oriented towards the lower fillets of the working and non-working faces across the rail head receive echo signals reflected from the lower and upper planes of the longitudinal crack. In the prototype, the reception of echo signals from the planes of a longitudinal horizontal crack is not provided.

2. Install the second electro-acoustic transducer emitting ultrasonic vibrations across the rail head towards the lower fillet of the idle face of the rail head, and the echo signals reflected from it are received. The prototype uses only one EAA, oriented across the rail towards the lower fillet of the rail head from the side of the working face. It is mainly intended for evaluating acoustic contact and the position of the system relative to the longitudinal axis of the rail.

3. The temporal positions of the echo signals on each of the EAFs, taken from the lower and upper surfaces of the crack, evaluate the depth of the crack. In the prototype, such an analysis is not provided.

4. The conclusion about the depth and degree of development of a longitudinal crack is made taking into account the echo signals reflected from the planes of the longitudinal horizontal crack and the amplitude level of the echo signals (amplitude envelopes) from the lower fillets of the rail head. The use of amplitude envelopes for analysis from two EAFs, fixing attenuation of signal amplitudes from the lower fillets of the head by a longitudinal crack in different sections, allows to more fully display the size and position of the crack in the head projection and along the length of the rail. This allows you to unambiguously assess the contour of the crack.

Assessment of the depth of the crack in the prototype is not considered, and the contour of the crack is determined only on the basis of the shadow method, almost up to half the surface (from the side of the working face) of the rolling of the rail. This does not satisfy the requirements of even domestic regulatory documentation (less stringent than European), which provides for the determination of parameters by dividing at least two gradations (“narrow” and “wide” crack).

5. The periods of emission of ultrasonic vibrations by the transducers are selected based on the speed of their movement and the required resolution for estimating the dimensions of a longitudinal crack, which reduces the number of measurements and the amount of stored information at low speeds of EAP movement without compromising the quality of work. At high speeds, due to the short path of ultrasonic vibrations in the cross section of the rail head, it is possible to obtain the required resolution of defect assessment. This factor is especially important for diagnostic systems that monitor rails in a wide range of scanning speeds (for example, from 0 to 80 km / h), where the total flow of information received from all measuring sensors is significant.

In the prototype, the question of the frequency of measurements is not considered.

Between the set of essential features of the proposed method and the achieved technical results, there is a causal relationship, namely:

- installation of the second EAP, reception of echoes reflected from the lower and upper planes of the crack, estimation of the depth and degree of development of the longitudinal crack based on the analysis of the amplitudes and temporal position of the echoes from the crack and the level of amplitudes of the echo from the lower corners of the rail head increases the reliability of the control of the rail head;

- the use for the detection and evaluation of longitudinal horizontal cracks of two EAPs oriented across the rail head can significantly reduce the travel time of ultrasonic vibrations in the section of the rail head and realize a higher frequency of probe pulses, and therefore, increase the speed of the EAP along the surface of the rail head and, in ultimately improves rail control performance;

- increasing the reliability and performance of the control of the rail head by using only two inclined EAFs greatly simplifies the implementation of the method, and the choice of the periods of radiation of ultrasonic vibrations by the transducers based on the speed of movement and the required resolution for evaluating longitudinal cracks allows you to reduce the number of measurements and the amount of information stored at low speeds EAP without deterioration of reliability and also simplifies the implementation of the method.

In FIG. 1 shows the process of sequential sounding of a longitudinal crack in three projections:

where 1 is the head of the controlled rail with working (G - gauge corner) and non-working (F - field corner) faces;

2- longitudinal crack;

3 - the first electro-acoustic transducer (EAP);

4- second EAP;

I, II, III, IV and V are sections of the rail head during the development of a longitudinal crack and different positions of the EAA as they move along the rolling surface of the rail head.

In FIG. 2 - shows the formation of signals on the EAA 3 and 4 at different degrees of development of a longitudinal crack in the rail head, corresponding to sections I, II, III, IV and V in FIG. 1, where:

5 and 6 are the axes of the radiation pattern of the EAA 3 and 4, aimed at the lower angles of the working (AG - angle gauge) and non-working (AF - angle field) faces of the rail heads, respectively;

7 and 8 - the extreme rays of the ultrasound beams of the EAP 3 and 4.

FIG. 3 - echoes on a scan of type A from different points of the rail head with a longitudinal crack, where:

in FIG. 3A - echo signals on the EAP 3 from the side of the working face:

- CG (crack, gauge corner) from the lower plane of the crack, obtained by re-reflection of ultrasound rays from the lower plane of the head from the side of the working face to the lower plane of the crack and back through the lower plane of the head to the EAT;

AG - signal from the lower fillet of the head;

TCG - (top of crack, gauge corner) signal from the upper plane of the crack and the acute angle between the crack and the upper fillet of the rail head;

TS - (top of crack) a possible echo signal from the plane of the crack, directly under the EAP 3 (obtained by direct, without re-reflection, ultrasound beam).

In FIG. 3b signals to the EAP 4 from the side of the idle side of the rail head:

AF - echo from the bottom angle of the head;

TCF - (top of crack, field corner) signal from the upper plane of a developed longitudinal crack intersecting the entire width of the head, and the acute angle between the crack and the upper fillet of the rail head from the side of the non-working side.

TS - a possible echo signal from the plane of the crack directly under the EAA 4.

In FIG. 3 a and b pos. 9 and 10 - threshold levels for fixing the level of the amplitude envelopes of the signals from the lower fillets of the head on the EAA 3 and 4, respectively;

11 - zone of temporary selection (gating impulse).

In FIG. 4 shows the behavior of the amplitude envelopes of the echo signals from the lower fillets of the rail head when scanning a longitudinal crack 2 shown in FIG. 1, where:

12 and 13 are the amplitude envelopes of the signals from the EAA 3 and 4, respectively, formed by the values of the amplitudes of the signals from the lower fillets of the rail head as the EAA moves along the rail;

ΔL3 and ΔL ₄ are the lengths of the sections of decreasing envelope signal levels from the lower fillets of the head below threshold levels 9 and 10 (see Fig. 3), in the process of moving the EAT along the rail using the example of a longitudinal crack 2 shown in FIG. 1.

In FIG. 5 - examples of real transverse cracks of different degrees of development in the rail head, discovered by the proposed method:

a - the crack does not cross the longitudinal axis of the rail; b - the crack crosses the longitudinal axis; c - a crack crosses the entire width of the rail head.

In FIG. 6 shows: FIG. 6a shows signals on a scan of type B obtained by scanning a rail head of type P 65 with a real longitudinal crack of a very complex configuration; on Fig 6b - the same signals on the scan type A (by analogy with Fig. 3).

Consider the possibility of implementing the proposed method.

From the rolling surface of rail 1 with a longitudinal crack 2 symmetrically to the longitudinal axis of the rail, using inclined EAFs 3 and 4 across the rail head, ultrasonic vibrations are emitted in the direction of the lower fillets of the rail head of the working (AG - angle gauge) and non-working (AF - angle field) faces of the rail head respectively (Fig. 1). Echoes reflected from the lower fillets are received by the corresponding EAPs 3 and 4. In the absence of a longitudinal crack 2, the ultrasound rays reflected from the lower corners AG and AF of the rail head 1 confidently generate echoes of approximately the same amplitude (see Fig. 3). If there is a crack 2 in the initial stage of development (in this case, from 5 to 10 mm), part of the rays of the ultrasonic beam of the EAA 3, reflecting from the lower plane of the head from the side of the working face, fall onto the plane of crack 2 from the bottom (Fig. 2 - I). Due to reflections from the lower plane of the crack 2 and, mainly, from the corner reflector formed by the plane of the crack and the side surface of the rail head, a confident echo signal CG (see Fig. 3a) is generated on the EAP 3. By the temporary position of this echo signal CG relative to the probe pulse EP (emission pulse) it is possible to very accurately determine the depth of the crack nucleus relative to the rolling surface (Fig. 3a).

As the longitudinal crack 2 develops, the crack plane to a greater extent overlaps the ultrasonic beam from the EAF 3 (Fig. 2 - II). In this case, the signal from the lower plane of the crack 2 still arrives at the EAA 3, but the amplitude of the echo signal AG already becomes lower than the threshold level 9 (Fig. 3a). At the same time, due to reflections from the upper plane of the crack, signals appear from an angular reflector formed by the plane of the crack and the upper fillet of the rail head: TCG signal (see Fig. 2 - II). Due to the predominant orientation of a typical longitudinal crack (see Fig. 1, 2 and 6) with a certain inclination relative to the horizontal (see Fig. 5), the TC echo can also appear (Fig. 3a) from the plane of the crack formed by a straight line (without multiple reflections) by a beam. The temporary position of which on a Type A scan allows you to determine the depth of the crack plane inside the rail head (Fig. 2 - III).

With the further development of crack 2, the ultrasonic beam is completely blocked by the plane of the crack, and the signals AG and CG from the lower angle of the rail head and the lower plane of the crack cease to be received by the EAP 3. Only the signals from the upper plane of the crack TCG and TS will be present on the EAP 3 (Fig. 3a )

As the crack 2 propagates to the side of the non-working face F, the echo AF from the lower corner of the head from the side of the non-working face is interrupted (Fig. 2 - IV). Subsequently, as the crack crosses the entire width of the head, a TCF signal appears on the EAF 4 from the upper plane of the developed longitudinal crack and the acute angle between the crack and the upper fillet of the rail head from the side that is not working. It is also possible that a vehicle echo from the crack plane formed by a direct beam appears on the EAP 4 from the crack’s temporary position, which can further clarify the depth of crack 2.

As you know, angular reflectors at any location (hydroacoustic, radio and acoustic, for this case) are the most effective reflectors that form clear echo signals. Therefore, the echo signals from the corner reflectors formed by the upper plane of the crack and the wall of the working and non-working fillets (echo signals CG and CF in Fig. 2) will be clearly and unambiguously recorded by EAPs 3 and 4.

Since the temporary positions of the expected signals considered above, with a tolerance for wear of the rail head and the probable position of the desired crack, can be preliminarily calculated, for their reception it is possible to form pre-calculated time zones 11 (zones of temporary selection) taking into account the type and geometry of the rails (on Fig. 3 is shown only for the CG signal, for the remaining signals on the EAP 3 and 4 - similarly).

The considered signal generation options for different degrees of longitudinal crack development can be summarized in table

As can be seen from FIG. 1-3 and tables, based on a joint analysis of signals from the lower fillets of the rail head and echo signals generated from crack 2, we can unambiguously judge the degree of development of the desired crack along the width of rail head 1. In each of the cases under consideration (sections I, II, III, IV and V) the defect is fixed by three to four signals and their combinations are not repeated, which indicates the possibility of recognizing cracks of varying degrees of development with high reliability. In this description, the considered sections I - V are given to simplify the understanding of the essence of the method. In practical implementation, this analysis can be performed in each radiation cycle - the reception of ultrasonic signals.

Additionally, the length of the cracks along the rail length can be determined by the duration ΔL3 and ΔL4 of the disappearance (the amplitudes U _{AS of the} echo signals AG and AF from the lower fillets of the head below the threshold levels 9 and 10 in Fig. 3) by analyzing the amplitude envelopes 12 and 13 on the EAA 3 and 4, respectively (Fig. 4).

Obviously, from the temporary positions of the echo signals from the lower surface (CG) and from the upper plane (TCG, TC3 TC4 and TCF) of crack 2, at a known propagation velocity of ultrasonic vibrations in the metal of rail 1, it is possible to unambiguously determine the depth occurrence of cracks relative to the rolling surface. Moreover, at almost any stage of development. Thus, by analyzing the signals received during the sequential movement of the transducers along the rail with a certain speed, it is possible to obtain a complete image of a longitudinal crack in the rail head.

Due to the combination of essential features of the proposed method (additional radiation of ultrasonic vibrations across the head towards the lower angle of the idle face of the rail head, receiving echo signals from the lower and upper plane of the crack, a joint analysis of the echo signals from the crack and the envelope of signals from the lower corners of the head), it is possible not only to obtain an image of a longitudinal crack, but also to classify them according to the degree of danger in accordance with the current regulatory documents of European countries [13] and JSC Russian Railways [1] (see the bottom two lines of the table .).

The last two columns of the table show that by the presence of echo signals from the lower corners of the rail head, it is possible to control the stability of the acoustic contact under the EAP system 3 and 4 (there is contact / there is no contact).

It should be noted that when implementing the proposed method, an inclined input (depending on the type and configuration of the rail head, at an angle from 30 ° to 70 °) is used and shear (transverse) ultrasonic vibrations with a propagation velocity in the metal of 3.26 mm / μs are emitted . In this case, the temporary position of the echo signal from the upper plane of the crack, under the conditions considered above for a direct EAF (h = 12 mm), is already more than 9.0 μs. This provides a clear time separation of the echo signal from the crack and the probe pulse, providing the ability to measure the time interval between them.

The threshold levels 9 and 10 (Fig. 3) are selected so that the duration of the disappearance of the echo signals from the lower corners when hitting a crack (signal amplitudes AG and AF below the thresholds) when displaying control signals on a scan of type B [12] practically correspond to the length longitudinal horizontal cracks L ₂ in scanned sections (see Fig. 6). With the predominant development of longitudinal cracks from the side of the working face of the rail head, as a rule, there is a longer duration of signal reduction from the lower angle to the EAA 3 than to the EAA 4. The condition ΔL ₃ ≥ ΔL _{4 is} almost always satisfied (Fig. 4), which can be an additional sign of a longitudinal horizontal crack.

In FIG. 1, the configuration of crack 2 is given as an example. In practice, the propagation of a crack from the point of damage on the working fillet can occur not only in one direction, but also in the other, or, simultaneously, on both sides along the rail from the starting point. The principle of detection and evaluation of crack parameters does not change and corresponds to the essential features of the proposed method.

Signal processing, both by detecting and evaluating cracks, and by observing the quality of control according to the above sequence (algorithm), can be carried out both visually and in an automated mode. In high-speed control, the automated version on the on-board computer and the presentation of signals on a scan of type B are preferable (Fig. 6, for example, the image of signals from only one EAA 3 is shown). As you can see, even analyzing the signals from only one EAA, you can get a visual representation of a very complex longitudinal crack in the rail head: the length L _{2 of the} desired crack 2 (Fig. 1) in the scanned section of the EAA 3; temporary positions of the echo signals from (see Fig. 6b) the lower plane of the crack from below CG and the upper plane of the crack TCG and TC. According to these temporary positions, it is possible to trace the development of the crack in depth in each section (in Fig. 6b - conditional sections 3, 4, 5 and 6). An analysis of a similar scan of type B on an EAT 4 on a scan of type B (not shown in Fig.), Also synchronized with the speed of the EAP, allows you to supplement the information about the defect parameters in the part of the head from the non-working side of the rail head.

As can be seen from FIG. 3, and calculations using well-known expressions, the temporal positions of the farthest signals from the probe pulse in the considered method do not exceed 50 μs. With a double safety factor, the period T of the radiation of the probe pulses can be taken equal to 100 μs, which corresponds to the frequency F of the transmissions of the probe pulses equal to F = 1 / T = 10 kHz. At such a parcel frequency, even a scan speed of 120 km / h currently unattained, the sounding interval of the rail section will occur every 3.3 mm, which is quite sufficient for reliable detection of longitudinal cracks and contouring of their area (formation of a defect image). Of course, this contributes to a significant increase in rail monitoring performance. At the same time, at lower scanning speeds, without loss of information, it is possible to reduce the frequency of sending probe pulses simultaneously with a change in the scanning speed. This will reduce the flow of processed information while reducing the speed of movement of the converters.

When implementing the method, it is possible to use EAP both on the basis of piezoelectric transducers and on the basis of electromagnetic-acoustic (EMA) excitation of ultrasonic vibrations.

The placement of the EAA on the rolling surface relative to the longitudinal axis of the rail head can be performed in different ways:

- in one cross section of the head, when each inclined EAF is on the same side from the longitudinal plane of the head as the lower fillet of the rail head voiced by it (as shown in Figs. 1 and 2);

- in one cross section of the head, when the inclined EAF is located on the opposite side from the longitudinal plane of the head, relative to the lower angle of the rail head sounded by it. However, in this case, acoustic interference caused by the mutual influence of the EAP on each other is possible;

- in two sections mixed relative to each other in the longitudinal direction of the rail, implementing the first or second method. With a known distance between the EAA 3 and EAA 4 along the rail, and the known speed of their joint movement, signal processing (analysis of the signals according to the above algorithms) must be carried out by bringing the analyzed signals to a single section of the monitored rail.

The proposed method for detecting longitudinal cracks, as noted above, is mainly aimed at detecting and evaluating parameters of cracks having a predominantly horizontal orientation (with some inclination relative to the horizontal). The author at the time of preparation of the application does not know the methods that solve this problem. At the same time, the proposed method allows to detect longitudinal cracks of vertical orientation (defects of code 31 according to [1]), the detection of which is known to technical solutions (see, in particular, [14]), thereby complementing and expanding their capabilities. When a vertical crack is located in the lateral part of the rail head from the side of the working face, the amplitude of the echo signal AG from the lower angle from the side of the working face decreases, but the amplitude of the AF signal from the side of the non-working side remains unchanged (close to maximum). And when there is a vertical crack in the lateral part of the head from the side of the non-working side - vice versa.

Thus, the claimed method allows you to:

- detect sections of rails with longitudinal internal cracks;

- determine the size of the cracks, both in width and along the rail head;

- determine the depth of longitudinal cracks of horizontal orientation;

- evaluate the quality of acoustic contact under the transducers;

- increase the frequency of sendings of probe pulses, helping to increase the speed of control.

The method can be implemented, allows to increase the reliability and performance of the control while simplifying the implementation of the technical solution in practice. Reliable detection of the most massive longitudinal cracks in the rail head with the possibility of classification according to size (in plan) and depth, allows timely measures to be taken to eliminate them and prevent catastrophic consequences in rail transport.

Sources of information

1. Instruction “Rail defects. Classification, catalog and parameters of defective and acute defective rails ”Approved. Russian Railways by order No. 2499 r of 10.23.2014. - 140 p.

2. US 4700574.

3. US 6055862.

4. US 5777891.

5. WO 2009140446.

6. RU 2060493.

7. RU 2184374.

8. RU 2308027.

9. RU 2545493.

10. RU 2613574.

11. RU 2652511.

12. Markov A.A., Kuznetsova E.A. Defectoscopy of rails. Signal formation and analysis. Book 1. Basics. - SPb .: KultIn-formPress, 2010.290 s.

13. S. L. Grassie. International Railway Journal, 2001, No. 1, p. 13-17.

14. US 20130152691.

Claims (2)

1. The method of ultrasonic detection of longitudinal cracks in the head of the rail, which consists in the fact that the first electroacoustic transducer is installed on the rail surface, oriented reprogramming the rail head to the lower fillet of the head from the side of the working face with the possibility of receiving reflected signals from it, moving the transducer along the rail with at a certain speed, periodically emit ultrasonic vibrations, receive echoes reflected from the lower fillet of the head from the side of the working edge of the rail, lysing the amplitude of the received signals, characterized in that echoes reflected from the lower and upper planes of the longitudinal crack are received, the second electro-acoustic transducer is mirrored first, but directed to the lower fillet from the side of the inoperative rail head and functioning similarly to the first, together with the first second transducer move along the rail, an assessment of the depth and degree of development of a longitudinal crack is carried out on the basis of an analysis of the amplitudes and temporary position of the the echoes coming from the crack and the amplitude levels of the echoes from the lower fillets of the rail head that are arriving as the transducers move. 2. The method of ultrasonic detection of longitudinal cracks in the rail head according to claim 1, characterized in that the periods of emission of ultrasonic vibrations by the transducers are selected based on the speed of their movement and the required resolution for estimating the size of longitudinal cracks.

Images