RU2586090C1 - Method for magnetic inspection of weld joints of rails - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: invention relates to investigation of materials using magnetic appliances, particularly measuring changes of value of magnetic flux during change of nominal cross-section or metal structure with ferromagnetic properties. Method for magnetic control of rail weld joints consists in that flaw detection means is fitted with a device which generates a magnetic field in rail, flaw detection means is moved and magnetic field variation in rail is detected with a sensor, sliding along surface of rail, welded joint areas are detected, their coordinates are stored in diagnostic map of track section, while further detecting and storing forms of signals from welded joint areas, comparing with corresponding signals of previous measurements and based on said comparisons, a decision is made on detection and development of defects in welded joints of rails.

EFFECT: technical result is improvement of reliability of detecting and efficiency of control of rail welded joints.

1 cl, 5 dwg

Description

The invention relates to the field of research of materials using magnetic means, in particular, fixing changes in the magnitude of the magnetic flux when changing the nominal cross section or structure of a metal with ferromagnetic properties. The invention can be used in high-speed defectoscopy of railway rails to detect defects in welded joints of rails. A powerful magnetic flux formed by the magnetizing system in the rail, encountering an obstacle (a transverse crack, a zone of low magnetic permeability) flows around this obstacle and partially exits. Near this zone, an additional magnetic field appears on the surface, which is detected by the sensor.

In modern rail tracks, welding is the main way to form continuous jointless lashes (in 2014 already 70% of the main tracks of Russian Railways are jointless). About 5 million welded rail joints are operated on Russian railways. Basically, according to the strength characteristics, the zone of the welded joint differs little from the zone of the base metal of the rail. However, violations of the welding technology, the appearance of fatigue discontinuities sometimes lead to kinks in rails in the area of welded joints. In the last 10 years, up to 35% of rail bends on the railways of Russian Railways occur due to defects in the welded joint zone.

In order to timely detect defects of fatigue origin, welded joints are periodically (every 2 years of operation) monitored by the ultrasonic echo method in accordance with the regulatory technical document (NTD) of Russian Railways [1]. The control is carried out using an ultrasonic transducer with an input angle of 50 °, performing manual scanning in the heat-affected zone (± 200 mm from the center of the seam) with a scan step of 3 mm around the entire perimeter of the seam. Currently, this technology is the most time-consuming and low-tech operation of non-destructive testing of the rail track. Huge labor and financial resources are spent annually on the control of welded rail joints. control performance by this technology is only 40 joints per shift [2]. In practice, the requirements of technical documentation are often violated, and the reliability of control of welded joints of rails remains unacceptably low.

In order to mechanize the process and record the results of the control, a method and device for ultrasonic testing of welded joints of rails according to the patent [3] is proposed, which include the installation of several electro-acoustic transducers on different surfaces of the rail in the heat-affected zone, probing the entire cross section of the welded joint, processing together the echo signals and the calculation of the spatial position of the defect. The method allows several times to increase the control performance and objectively register all signals received from the welded joint area of the rail. The known method is the mechanized control of the welded joint, requires the participation of two operators (to move the device from joint to joint and quick removal of the device when the train is approaching) and has insufficient cost-effectiveness and control performance.

In addition, ultrasonic methods do not provide high-quality control of the upper part of the rail head, the zone of maximum load on the rail section, including in the welded joint zone. The current scientific and technical documentation of control of welded joints of rails (see, for example, [1]) allows the possibility of skipping defects at a depth of 8 mm from the rolling surface of the rail head. At the same time, contact fatigue cracks of code 26.3 and irregularities on the rolling surface of code 46.3 [4] begin to develop in the near-surface zone (head crushing due to uneven mechanical properties of the metal at the welded joint site) (see pp. 226-258 [ 5]).

In the case of continuous ultrasonic testing of rails with the help of inspection trolleys simultaneously with the control of the base metal zone, the zones of the welded joints of the rails are also controlled [5]. However, they are characterized by the above disadvantages of the ultrasonic method. As a rule, in the absence of internal defects in the welding zone, the weld joint is not localized on the defectogram of ultrasonic inspection (there are no echo signals).

On modern railways, high-speed flaw detection vehicles are widely used: cars, railcars, etc., which make it possible to occupy rail tracks for technological procedures to the least extent. High-speed flaw detection involves the use of appropriate methods, mainly ultrasonic and magnetic. Ultrasonic (ultrasound) methods allow a fairly detailed study of the internal structure of the rail in order to detect defects, but they have a number of problems with the application, in particular, ensuring reliable acoustic contact of electro-acoustic transducers (emitters and receivers) with the rail. As a result, the signals even from a stable reflector (for example, a bolt hole) during periodic monitoring are repeated with a low correlation coefficient and cannot be used to monitor the state of the controlled object (rail).

Magnetic (magnetodynamic) methods (MD) are easier to use, less dependent on external conditions. The advantages of magnetodynamic methods include: the ability to work in all climatic zones under any weather conditions; high reliability of measurements, providing good repeatability of the results and the ability to monitor the development of individual defects; the ability to work at any vehicle speed.

However, magnetic methods make it possible to detect rail defects only in the surface and near-surface areas of the rail head. When using modern magnetization systems with electromagnets on the axles of wheelsets of a special carriage (inductor) carriage, the depth of occurrence of the detected MDs by the method of defects can reach 15 mm [6].

In addition, the MD method allows you to solve the problem of relative navigation by tying the position of flaw detectors to the structural elements of the rail track: bolted and welded joints, butt plates, turnouts (and even due to the fixation of the scattering field to the rail linings) and other anomalous objects that can be detected by the method. Differentiation of signals from defects and structural elements can be solved by comparing the coordinates of the detected anomalous objects with the coordinates of the structural elements stored previously in the diagnostic map of the rail track section.

Magnetic methods for flaw detection of rails have long been known, for example [7], and consist in the fact that a device that creates a magnetic field in the rail is installed on the flaw detector, the flaw detector is moved and the magnetic field in the rail is fixed with a sensor sliding on the rail surface, and anomalous in amplitude is detected signals that decide to detect defects in the rail. This method involves the exclusion from consideration of welded joints of rails. The disadvantage of this method is the low reliability of detection of defective sections of the rail track, in particular in the area of welded joints of the rails.

A known method of magnetic detection of defects in rails [8], which consists in the fact that a device that creates a magnetic field in the rail is installed on the flaw detector, the flaw detector is moved and the magnetic field in the rail is fixed with a sensor sliding on the rail surface, the signals of which are compared with reference signals defects in a correlation way.

The disadvantage of this method is the difficulty of creating a library of reference signals, which today does not exist due to the features of the magnetic method noted above.

Closest to the claimed is a method for diagnosing a rail track [9], which consists in installing a device that creates a magnetic field in the rail on the flaw detector, moving the flaw detector and fixing the magnetic field in the rail with a sensor sliding on the rail surface, detect signals from structural elements of the rail track and defects, save their coordinates in the diagnostic map of the section of the rail track. Moreover, as follows from the description of the known method, the signals from the zones of welded joints of the rails, as well as other structural elements of the track, can be selected both manually and automatically.

The disadvantage of this method is the low detecting ability of defects in the area of welded joints of rails. This drawback is due to the fact that, both in this and in other methods known to the authors, the informational capabilities of the magnetic control method are not fully used.

The problem solved by the claimed method is to increase the reliability and performance of the control of welded joints of rails using the magnetic control method.

To solve the problem, a device that creates a magnetic field in the rail is installed on the flaw detector, the flaw detector is moved and magnetic field changes are recorded in the rail by a sensor sliding on the rail surface, weld joints are detected, their coordinates are saved in the diagnostic map of the rail track section, and additionally fixed and save waveforms from the zones of the welded joints, compare them with the corresponding signals of previous measurements, and based on these comparisons accept dissolved solution and the detection of the development of defects in the welded joints of the rails.

Of all the structural elements that are fixed during scanning of the rail track by the magnetic control method, implemented by a flaw detector (for example, a flaw detector car), the most stable and repetitive from passage to passage are the signals from the zones of the welded joints of the rails. Signals from bolted joints can change over time due to temperature changes in butt gaps, signals from turnouts depend on the direction of movement and, accordingly, on the position of the wit translation. Only signals from the zones of the welded joint, while the joints are unchanged, remain stable for a considerable period of time.

According to specialists from the Welding Department of VNIIZHT, in the general case, the metal structure of the welded joint of the rails consists of six sections: a seam; incomplete melting zone; recrystallization zone (large grain zone); recrystallization zone (fine grain zone); zone of incomplete normalization and base metal [10]. Thus, in the process of moving (driving through) the flaw detector above the welded joint, the magnetic field sensor travels from one zone of the base metal to another nine zones with different structures, usually referred to as the "heat affected zone" or the welded joint zone. The total size (length) of this zone depends on the welding technology used (pulsating or continuous reflow) and can be up to 80 mm. Naturally, the magnetic permeability of the component parts of the weld zone is different, which ensures the formation of a characteristic signal (response) of the magnetic control method during scanning.

As noted by the authors of this application, the signal shape of the MD method can significantly change when the state of the welded joint changes during operation: the generation and development of internal cracks in code 26.3, the appearance of irregularities on the rolling surface of code 46.3 (wrinkling of the head due to uneven mechanical properties of the metal in the welded place joint), metal chipping at the working fillet (defect code 10 or 11) [5] and mechanical damage to the rail head. This regularity is confirmed by observations of the state of the welded joints of the rails for a sufficiently long period of time according to the signals of the magnetic method and additional studies by alternative methods (visual inspection and ultrasound studies) of different sections of the rail track.

At the same time, a high frequency of signals from defect-free welded joints was noted both in amplitude and in the shape of the signals. Thus, it was found that the signals of the MD method can be used not only as indicators of anomalies in the rail, but also have information that allows us to obtain more detailed information about the state of the welded joint of the rails by the shape of the signal.

Significant differences of the proposed method are that:

Additionally measure and save waveforms from the weld zone by the magnetic method. The waveform contains information about the state of the weld, the change of which during periodic monitoring can be used to fix the nucleation and development of defects in the controlled area, to identify detected objects and evaluate their properties. In the claimed technical solution to the procedure, fixing and analyzing the shape of the signals from the zones of the welded joints are key.

In the prototype, signals are detected from structural elements of the rail track, including welded joints, in amplitude, for which only their coordinates are stored in the diagnostic map. Thus, the waveform is not considered.

The waveforms from the zones of the welded joints of the rails of the magnetic method are compared with the shape of the corresponding signals of the previous measurements and a decision is made on the detection and development of defects in the zone of the welded joints of the rails. As you know, the waveform - the dependence of the instantaneous value on time - allows you to evaluate many parameters of the oscillations. For automatic comparison and detection of changes in the form of current signals from the zones of welded joints with the previous ones, various analysis methods can be used: spectral, correlation, standard deviation of the amplitudes of the RMS signals and others.

In the prototype, the analysis of the stored signals of the magnetic method is performed only from the point of view of the navigation problem. The signals from a magnetic flaw detector are used only as reliable coordinate marks of the corresponding structural elements of the rail track, i.e. serve to solve the binary problem: is there or not the desired element at a given point. Moreover, in the prototype, the analysis of the waveform from the heat affected zone of the welded joint is not performed.

The inventive method is illustrated by the following graphic materials:

FIG. 1 - flaw detector (flaw car), where

1 - Flaw detector car.

2 - Sensor magnetic flaw detector.

3 - Track sensor (from the wheel).

4 - GPS equipment.

5 - Processing unit.

6 - Diagnostic card carrier.

7 - Rail.

8 - Magnetization coils.

FIG. 2 - Scheme of a magnetic flaw detector, where

9 - Magnetic flux.

10 - Welded joint of rails.

11 - Surface defect of the rail head.

12 - Internal rail defect.

13 - Sleeping lining.

FIG. 3 - Signals of the MD method from the welded joint of the rail:

A) Original;

B) In a month;

C) After a year;

14 - signal from the welded joint;

15 - signal from a defect in the welded joint.

FIG. 4 - Signals of the MD method from the zone of the welded joint of the rail:

A) Original;

B) In a month;

C) After a year;

16 - signal from the welded joint;

17 - signal from a defect in the welded joint zone.

FIG. 5 - Results of correlation signal processing of the MD method.

Consider the possibility of implementing the proposed method on the example of processing signals of the magnetic method in the area of welded joints of rails.

A flaw detector car (or any moving unit on a rail: a car, a car on a combined course 1, Fig. 1, is mounted on rails 7 and equipped with a magnetic flaw detector, which includes magnetization coils 8 mounted on the axles of a pair of wheels, a magnetic flaw detector 2 and processing unit 5. In addition, navigation equipment is installed in carriage 1: GPS equipment 4 (absolute coordinate sensor) and track sensor 3 connected to processing unit 5.

The navigation system of a flaw detector car, on the one hand, should provide the determination of navigation parameters at long distances (tens, hundreds of kilometers), and on the other hand, determine the coordinates of defective sections several millimeters in size with an accuracy of centimeters. These circumstances lead to the fact that the navigation system, in accordance with the known method adopted as a prototype, is built on a hierarchical basis. The absolute navigation subsystem GPS 4 generates data on the coordinates of the flaw detector car. These data are inaccurate (errors are meters), but they allow you to avoid gross binding errors. The relative navigation subsystem "from the wheel" 3 has recursive properties, i.e. it is able to accumulate errors, however, in short sections, its data is highly accurate and allows you to determine the distance from the selected marker to the point of interest.

The navigation system of a flaw detector car may contain subsystems that form appropriate markers, for example, associated with kilometer and picket posts, detected automatically or manually (not shown in the drawing). A more accurate navigation reference can be achieved using welded rail joints, located, as a rule, at a distance of 25 m from each other and detected by a magnetic flaw detector. Improving accuracy can also be achieved by rail linings, also detected by a magnetic flaw detector due to the scattering fields of the magnetization system of the flaw detector car.

In the process of movement of the flaw detector car, FIG. 2, the magnetization coils 8 excite a magnetic field 9, which is closed by metal structures: the wheel axis, the wheel itself, the rail, the wheel, the wheel axis, etc. The field penetrates the rail to a depth of 15-20 mm, as a result of the MD method, inhomogeneities are mainly detected: welded joints 10, surface 11 and subsurface defects 12 mainly in the rail head. The magnetic field of scattering covers the surrounding space, including structural elements, for example, sleeper pads 13. When the flaw car moves, the magnetic field is perceived by a sensor 2 sliding on the rolling surface of the rail. The received signals are digitized, analyzed in processing unit 5 and stored in the diagnostic card 6 with reference to the coordinates of the path.

As the sensor 2 of a magnetic flaw detector can be used induction, magnetoresistive, flux-gate transducers or Hall sensors. The measurement accuracy of these transducers is different and depends on the actual scanning speeds. In this case, the sensors 2 of the magnetic field of the induction type are located on the surface of the rail near the rear (relative to the direction of movement of the car) magnetic pole of the magnetization system (rear wheel of the induction trolley).

Signal processing of the MD method in the processing unit 5 consists, firstly, in the selection of signals from the zones of welded joints according to the amplitude of the signals and storing their coordinates in the diagnostic map 6 according to the data of the navigation system. Secondly, when such a signal is detected, the shape of the received signal is stored in the diagnostic card 6 in a certain vicinity of the anomalous amplitude. Taking into account the fact that, according to the current technical documentation, ultrasonic testing provides scanning at a distance of ± 200 mm from the center of the weld [1 and 5] for the purpose of diagnosing the welded joint of rails using the magnetic method, in order to preserve the unity of approaches (with a significant margin), we will accept a track section 400 mm long with analysis of the corresponding section of the defectogram.

If the measurements are not the first, then the diagnostic card stores the results of previous measurements. Thirdly, the forms of the current signals from the zones of the welded joints are compared with the previous ones using mathematical methods of analysis (spectral, correlation, or standard deviation of the amplitudes of the RMS signals), on the basis of which they decide on the detection and development of a defect in the area of the welded joint of the rail. The authors of the prior art do not know how to control welded joints of MD rails by the method of analyzing waveforms from heat affected zones using analysis methods.

In FIG. Figure 3 shows the actual form of the signals of the magnetic method 14 from the welded joint of the rails obtained in A) - primary measurements, B) - in a month and C) - in a year. Obviously, the signals 14 A) and B) are almost identical, which indicates good repeatability of measurements, and also that there were no significant changes in this welded joint in a month. Result C) has a shape distortion (an additional impulse 15 has appeared), i.e. about the possible development of a defect. Visual inspection of this joint and ultrasound studies using the existing technology [1 and 5] showed that a transverse crack of code 26.3 develops inside the welded joint.

In FIG. Figure 4 shows the actual form of the signals of the magnetic control method 16 from another welded joint of the rails, and after it the signal 17 - from the edge of the heat affected zone of the welded joint obtained in A) - primary measurements, B) - in a month and C) - in a year. Comparison of A) and B) again confirms the repeatability of the results for signals 16 and 17. Signal C) 17 has significant differences from A) and B), which indicates the possible development of a defect at the edge of the heat-affected zone. It is noteworthy that the signals A), B) and C) 16 directly from the welding are identical to each other and have not changed during the year. Visual inspection of this joint and ultrasound studies showed that from the existing irregularities on the surface of the rail head, the development of a transverse crack inside the head begins.

To automatically detect these changes in the signals of the MD method, various analysis methods can be used: spectral, correlation, and others. In order to avoid narrowing the claims, it is proposed that the specific analysis method not be indicated in the claims.

Correlation analysis involves the calculation of the correlation coefficient K between the signals S ₁ and S ₂ from the object received at different times:

The parameter w in (1) determines the half-width of the correlation window, and by maximizing the shift parameter s, the relative position of the responses from the object on the MD records is refined.

As an example, we consider the results of correlation processing of measurements (Fig. 5) obtained by the AVIKON-OZM [6] flaw detector complex [20] of welded joints made with a time interval of about a month. For all joints under consideration, with the exception of the fifth joint (in Fig. 5, the fifth point from the vertical axis of the graph), the correlation coefficient K is at the level of 0.8-0.95. The correlation coefficient of the fifth joint is 0.7. The choice of the threshold for the automatic determinant of the anomalous (defective) weld joint at the level of 0.75 allows us to state that weld joint No. 5 is anomalous and is suspected of developing a defect. Additional studies of this joint using current technology [1] allow us to make the final decision on the feasibility of operating this joint. In practice, the choice of the threshold is carried out according to the results of experimental studies.

In the examples given (Figs. 3, 4, and 5), due to the proposed method, it was possible to detect defective joints by a magnetic method and a defect in the weld zone (at the boundary of the heat-affected zone), without waiting for a periodic (with a period of 2 years) inspection of the weld zone junction according to current technology [1] and thereby prevent sudden breaks in rails under trains. In heavy and high-speed sections, the frequency of rail monitoring using mobile means of control reaches up to four times a month (weekly passage of sections of the track by a flaw detector car, equipped with, among other things, a magnetic flaw detector (channel)). Implementation of the proposed method will allow monitoring the condition of welded joints with a specified period, timely record the change in the state of the joint and take proactive measures to prevent fractures of rails. Moreover, the task of controlling welded joints is solved in the process of fulfilling the main task of the MD flaw detector - searching for defects in the rail head at significant scanning speeds (up to 120 km / h).

The stability and high repeatability of the signals from the zones of the welded joints of the rails for several years of their operation allows us to offer a kind of “passport” of the welded joint based on the images of the magnetic method, where, in addition to the signals from the joint, the name of the manufacturer, year and number of melting of the connected rails ; date, place (No. RSP), modes of welding, grinding and induction hardening of the joint; waypoint coordinates (direction number, way number, line, km, station, junction number at 100 m (at station) section). In this case, the most objective (independent of the operator entering the input data) parameter is precisely the waveform from the weld zone obtained by the MD method. Moreover, this form is a unique characteristic for each specific welded joint, as it is caused by many subtle nuances of the welding manufacturing process (see, for example, signals from the weld 14 in Fig. 3 and 16 in Fig. 4).

Thus, the claimed method can be implemented and can improve the reliability of detection of defects in welded joints of MD rails by the method by detecting a change in the state of the joint at an early stage of development of a potential defect. Scanning at significant speeds greatly improves the performance of weld inspection.

Information sources

1. Technological instructions for ultrasonic testing of welded joints of rails in rail welding enterprises and on the way. TI 07.42-2004.

2. Regulation on the system of non-destructive testing of rails and the operation of means of rail defectoscopy in the track facilities of railways of JSC Russian Railways. - Rasp. Russian Railways OJSC No. 2714r of December 27, 2012.

3. Patent RU No. 2309402.

4. Classification of rail defects. NTD / CPU 1,2,3-93, M .: Transport, 1993.

5. Markov A.A., Shpagin D.A. Ultrasonic flaw detection of rails. 2nd ed. reslave. and add. - St. Petersburg: Education-Culture, 2013. - 283 p.

6. Antipov A.G., Markov A.A. Assessment of the depth of detection of transverse cracks by the magnetodynamic method in rail inspection. // Defectoscopy (RAS), 2014, No. 8, p. 57-68.

7. Copyright certificate SU 57745.

8. Copyright certificate SU 1675146.

9. Patent RU 2521095.

10. Gudkov A.V., Nikolin A.I., Turbina L.A. and other. Contact welding of rails and heat treatment of welded joints of rails of modern production at rail welding enterprises of Russian Railways. // Welding and surfacing technologies in railway transport: Sat. scientific Tr. OJSC VNIIZHT / Ed. A.V. Gudkova. - M .: Intext, 2008 .-- 176 p.

Claims (1)

The method of magnetic control of welded joints of rails, which consists in installing a device that creates a magnetic field in the rail on a flaw detector, moving the flaw detector and recording changes in the magnetic field in the rail with a sensor sliding on the rail surface, detecting zones of welded joints, and saving their coordinates in diagnostic map of the rail track section, characterized in that it additionally fixes and stores waveforms from the zones of the welded joints, compares them with the corresponding signals Alami previous measurements and on the basis of these comparisons, decide about the discovery and development of defects in welded joints of rails.

Images