Rail Flaw Detection

A defect in a rail which will ultimately lead to the fracture or breakage of the rail is called a flaw. From the point of view of the ultimate consequence of the flaw resulting in a fracture, it is necessary to detect these flaws and take timely action to remove them. Rail flaws can be detected either by visual examination of the rail ends or by rail flaw detection equipment.

Visual examination of rail ends In this method, the joint is first opened after removing the fish plates. The rail ends are then cleaned using kerosene oil and visually examined in detail with the help of a magnifying glass for any hair crack, etc. White chalk is sometimes rubbed on the rail ends so as to identify the flaw clearly. A mirror is used to reflect light on the joint in case sufficient light is not available.

Ultrasonic rail flaw detectors Ultrasonic rail flaw detectors (USFDs) have been progressively used in recent years on Indian Railways for the detection of flaws. This method is also known as the non-destructive method of testing rails.

6.8.1 Theory of Ultrasonic Rail Flaw Detectors

Vibration waves of a frequency of more than 20,000 cycles per second are termed as ultrasonic waves. These waves have the property of being able to pass through materials and following the normal principles of light waves of refraction, reflection, and transmission. Whenever there is a change of medium, some of the ultrasonic energy gets reflected and the rest gets transmitted. The amount of energy reflected depends upon the physical properties of the two media. When travelling through steel, if these waves come across air either from the bottom of the steel or from any flaw inside the steel, the reflection is almost 100%. This property has been found most useful for detecting flaws in rails. Thus, when ultrasonic waves are fed from a location on a rail, they pass through the rail metal and are normally reflected only from the foot. However, if a discontinuity exists in the rail metal due to some flaw, the ultrasonic waves get reflected back from the location of the flaw, which can be picked up and the defect located.

Production of ultrasonic waves

There are several methods of producing ultrasonic energy. The most common and simple method of producing ultrasonic frequency is by using ‘crystal transducers',

which normally produce ultrasonic waves of a frequency of up to 15 MHz. The crystals generally used for this purpose are made either of quartz or of barium titanate, cut to special size, shape, and dimensions. These crystals have the peculiar property of changing dimensions and generating vibrations in a particular direction when an oscillating electric charge is applied to the crystal faces. Also, when these crystals are made to vibrate, they produce an oscillating electric current. The crystals, as such, have the potential of generating ultrasonic vibrations, as also of converting the waves received after reflection into electric current. They also possess reversible properties.

These crystals are housed in metal holders protected by superior quality Perspex and then termed probes. There are two types of probes used for ultrasonic testing.

Normal probe This probe consists of two semi-cylindrical thin crystals with a vertical separating layer through the crystals and Perspex. These probes transmit ultrasonic waves vertically downwards when put on the rail and are suitable for detecting horizontal or inclined flaws, including bolt hole cracks.

Angle probe In this probe, crystals are mounted on an angular surface capable of transmitting pulses at an outward angle, which may be forward or backward or both, with separate transmitting and receiving crystals. The waves emitted by these probes follow an inclined path and are suited to detect inclined and vertical defects.

Techniques of ultrasonic testing

A number of techniques have been used in ultrasonic testing to suit the design of different equipment. Some of these techniques are the following.

(a) Frequency modulation

(b) Pulse echo

(c) Transmission

(d) Resonance

(e) Acoustic range

Indian Railways uses the frequency modulation and pulse echo techniques and only these are discussed in detail here.

Frequency modulation technique In instruments utilizing frequency modulation ultrasonic waves are created with the help of a probe crystal and transmitted continuously into the rail at rapidly changing frequencies. It is necessary for the rail head to be wet to enable the ultrasonic waves to pass efficiently from the crystal to the rail. The waves that get reflected from the opposite face are received continuously by the crystal. There is interference between the transmitted waves and the reflected waves, which causes resonance. As the frequencies of the transmitted waves are changing constantly, such resonance takes place at regular intervals. When the position of reflection is changed due to a flaw in the metal, the resonance gets affected, which can be easily detected by the operator. Instruments manufactured on this principle such as the Audi-gauge are light, portable, and simple in mechanism. However, these instruments have certain limitations.

(a) Fine vertical cracks are not readily detected because the single vertical probe does not find any surface defect from which it can be reflected.

(b) Cracks wholly below bolt holes are also not detected, as the vibrations are interrupted by the hole.

Pulse echo system In the pulse echo technique, a pulsed ultrasonic beam of very high frequency is produced by a pulse generator and sent in to the rail. At the opposite face, the ultrasonic waves are reflected and the echo is picked up by the crystal transducers. A discontinuity or defect in the rail will also produce the echo. The time interval between the initial pulse and the arrival of the echoes is measured with the help of a cathode ray tube. There may be multiple reflections of the echo but the one arising due to a fault can easily be determined by its relative position and amplitude.

The more sophisticated types of instruments that are based on the pulse echo system are being manufactured by the firm Kraut Kramer at present.

6.8.2 Kraut Kramer Multi Probe Rail Testing Trolley

This is the most common type of equipment used on Indian Railways for detecting flaws in the rail Fig. 6.8. The equipment is fitted on a hand trolley that is carried on the rails. There are two probes: a normal probe and an angle probe, both act independently. The probe material used for the production of ultrasonic waves is barium titanate, which produces and transmits vertical ultrasonic waves of four megacycles frequency through the vertical probe and 2 mega cycles frequency through the angle probe. The cylindrical probe is mounted on a knuckle jointed holder frame and has a renewable bakelite wear plate at the base. As the probe is worked over the rail, the bakelite piece takes the wear completely. The height of the probe above the rail surface can be adjusted in the holder assembly. The normal probe is powerful enough to scan the entire rail depth for defects. It can detect longitudinal discontinuities in the head at the junction of either the web and the foot or the web and the head as well as cracks from bolt holes. It cannot, however, detect vertical cracks. The defects detected by the normal probe are represented on the oscilloscope screen in the form of firm echoes protruding from a base line. Ordinarily, two echoes are visible on the screen, the initial echo due to the partial reflection of the waves from the rail top and the back echo from the bottom of the rail. Any echo between the initial and the back echo with a corresponding reduction in the height of the back echo is termed a flaw echo and is indicative of a flaw. The position of the flaw can be known by reading the distance of this intermediate echo from the initial echo, which will be the distance of the flaw from the rail top.

The ultrasonic rail testing trolley can be used without any block protection, but one has to be vigilant about the movement of trains. Progress depends upon the experience and the efficiency of the operator. The work is quite strenuous in nature and a single operator cannot observe the screen continuously for a long time. Work can also not be done during the middle of the day in the summer months because the operator will not be able to pick up the signals clearly. On account of these limitations, the work progresses rather slowly and approximately 2-3 km of rail are covered per day with two operators.

Fig. 6.8 Ultrasonic rail testing trolley

6.8.3 Classification of Rail Flaws

Depending upon the nature and extent of internal flaws, traffic density, and speed on the section, the defects noticed by rail flaw detection methods have been classified into three major categories, i.e., IMR, REM, and OBS.

IMR defects A defect that is serious in nature and can lead to sudden failure is classified as IMR. Immediately after detection, clamped fish plates should be provided for the defective portions and a speed restriction of 30 kmph imposed till the IMR rail is removed by a sound-tested rail piece of not less than 6 m. A watchman should be posted till the clamped fish plates are provided to avoid any mishaps. IMR stands for immediate removal.

REM defects These are the type of defects that warrant early removal of the rail from the track. These defects are marked with red paint. REM stands for remove.

OBS defects These are defects that are not so serious. The rail need not be removed in such cases but should be kept under observation. These defects are marked with yellow paint. OBS stands for observe. OBS defects have been further classified as OBS (E) and OBS (B). An OBS defect located within the fish-plated zone is designated OBS (E). Similarly, an OBS defect on major bridges and up to 100 m on their approaches is designated OBS (B).

In the case of the need-based concept of ultrasonic testing (explained in Section 6.8.6), there are only two types of defects, IMR and OBS. However, if the defects occur at welded joints, they are called IMR (R) and OBS (W) defects.

The actions to be taken upon the detection of defective rails are summarized in Table 6.10.

Table 6.10 Actions to be taken upon the detection of defective rails

Classification of defects

Action required

Speed restriction to be imposed if the required action is delayed


(red - 3 stars)

Immediate replacement (not later than 3 days)

Impose 30 km/h and depute a watchman till defective part replaced


(Red - 2 stars)

Replace within 15 days

Impose 30 km/h if not replaced within 15 days

OBS (E) (red - 1 star)

Replace or end crop within 15 days

Impose 30 km/h if not replaced within 15 days

OBS (B) (Red - 1 star)

Replace within 15 days

Impose 30 km/h if not replaced within 15 days

In the case of OBS rails other than OBS (E) and OBS (D), permanent way inspector (PWI) should observe each OBS location with a magnifying glass and duly record his observations once a month, to see if the crack has developed any further, in which case the same action as for REM defects should be taken. The PWI should maintain sleepers, fittings, and the ballast at such locations in sound condition. The assistant engineer should also test-check some of the OBS locations and record his observations during his monthly push trolley inspection of each section.

6.8.4 Ultrasonic Rail Flaw Testing Car

As the portable ultrasonic rail flaw detector can test only 2-3 km of rails everyday, advanced railways such as German Railways have developed a rail testing car that can test a much longer length of track much more effectively. The test car tests the track at a speed of 30 kmph and each of the two rails is tested ultrasonically by means of five probes (0°, ±35°, ±70°) and an airborne sound assembly. The test results are recorded photographically on a tape, a subsequent examination of which can reveal all the flaws. The flaws can then be properly classified. Their position in the track can also be pinpointed with respect to kilometrage to an accuracy of 1 m. This special car tests the track during the day and covers 100-200 km per day. The car covers approximately 20,000 km of track every year on German Railways. The average cost of testing is Rs 300-500 per km of the track. With the present trend in increase in speed, the need for ultrasonic inspection of rails is felt all the more to avoid hazards due to rail fractures, and in this context, the use of the test car on Indian Railways is considered a technical and operational necessity.

6.8.5 Self-propelled Ultrasonic Rail Testing Car

Indian Railways has recently procured a self-propelled ultrasonic rail testing (SPURT) car of make MATRIX-VUR-404 from Sa Matrix Industries, Paris, at a total cost of approximately Rs 25 million, inclusive of ancillary equipment. The MATRIX car is capable of detecting, measuring, recording, and simultaneously analysing the internal defects of rails using a non-destructive method of rail flaw detection. The rail testing car has been designed for simultaneous examination of

rails, points, and crossings at a maximum speed of 40 kmph. It consists of the following parts.

(a) An ultrasonic detection lorry

(b) An electronic unit including all circuit transmitters, receivers, selectors, and auxiliaries

(c) Two multi-track recorders installed at either end of the car

(d) A real-time automatic defect analyser

The various components of rail flaw detecting equipment are shown in the schematic diagram in Fig. 6.9. Testing is done using five probes at different angles, which are able to detect rail flaws.

Fig. 6.9 Various components of ultrasonic rail flaw detecting equipment

The SPURT car can detect most rail defects that normally develop under traffic during service. The type of defect, its size, and its position in the rail section is automatically determined. The SPURT car is able to screen the rail section completely in the web and almost completely in the head and the zone of the foot below the web. The flange of the foot and the top corners of the head, however, are not screened. The defects recorded are automatically analysed. The results are given in a synthesized form in a prescribed manner.

It is expected that the SPURT car will be used intensively on Indian Railways and that it will be possible to control rail fractures/failures on Indian Railways to a considerable extent.

6.8.6 Need-based Concept of USFD Testing

Indian Railways has decided to introduce on its system a need-based concept of USFD testing based on a Russian concept. As per the present policy, need-based rail inspection is being progressively introduced on A (not covered by the SPURT

car), B, and D routes on Indian Railways. The introduction of this concept will require changing the present classification of defects, frequency of inspection, detection equipment, organization, etc.

The following are the important features of the need-based scheme of USFD testing.

Traffic density and periodicity In the need-based concept, the stipulated frequency of ultrasonic inspection is one after the passage of every 8 GMT, with periodicity varying from 2 to 6 months, depending on the sectional GMT.

Important related parameters The system has been evolved based on the consideration of two important related parameters: permissible condemning defects size and inspection frequency. Other important factors such as microstructure of rail steel and nature and orientation of cracks, have not been taken into consideration.

Defect size Under the concept, a defect size of 12 mm and above in the head and 5 mm and above in the web junction is taken into account for the classification of defects. Therefore, defects sized below 12 mm and 5 mm are allowed to continue in the tracks as unclassified defects. Further, the same size of artificial flaw, i.e., 12 mm, is considered when classifying the weld defect too. Attenuation or absorption of ultrasonic energy is supposed to be more in SKV/AT welded joints due to coarse grain, whereas absorption of energy is less in rail steel. (SKV is a German word meaning short preheating. AT denotes alumino thermit.) Obviously, the specified 60% peak height will not be available in the same grain setting in the case of welds.

Frequency of testing of rails and welds An inspection frequency of 8 GMT has been prescribed for the need-based concept of ultrasonic testing. A higher inspection frequency may be fixed depending upon the incidence of defects. In view of this, whenever the defect generation rate (failures in service and defects detected during USFD inspection) exceeds 1 per km between successive tests in a stretch, the inspection frequency should be doubled in that particular stretch. When calculating the defect generation rate, only rail defects (IMR) or fractures with an apparent origin other than the bolt area and detectable by USFD should be considered.

Testing of rails After the initial testing of rails in the rail manufacturing plant, the first re-test normally need not be done until the rails have covered 15% of their service life in GMT as given in Table 6.11. For rails rolled in April 1999 and later, the test-free period will be 25% instead of 15%.

Table 6.11 Service life of rails


Rail section

Assessed GMT service life for T-12-72 UTS rails

Assessed GMT service life for T-12-90 UTS rails


60 kg



52 kg



90 R




75 R



60 R



Whenever rails are not tested in the rail manufacturing plant, the test-free period will be applicable, and rail testing will be done at the periodic interval given in Table 6.12 right from the day of its laying in the field. This table gives the frequency of ultrasonic testing after the passage of 8 GMT, subject to a maximum interval of one year.

Table 6.12 Frequency of ultrasonic testing for all BG routes


Testing frequency (months)

Up to 8


8 < GMT < 12


12 GMT < 16


16 < GMT < 24


24 GMT < 40


> 40


In the need-based concept, the actions suggested in Table 6.13 are taken when defective rails are detected.

Table 6.13 Action to be taken after detection of rail/weld defects


Painting on both faces of the web

Action to be taken

Interim action


Three crosses

The flawed portion

PWI/USFD (ultrasonic flaw


made with

should be replaced

detecting officer) shall impose

red paint

by a sound-tested rail piece of not less than 6 m length within three days of detection

speed restriction of 30 kmph immediately and this to be continued till the flawed rail weld is replaced. He should communicate the location of the flaw to the sectional PWI, who shall ensure that a clamped, joggled fish plate is provided within 24 hrs.


One cross

Rail/weld to be

• PWI/USFD to advise


made with

provided with

sectional PWI within 24 hrs

red paint

clamped, joggled fish plate within three days. PWI/USFD to specifically record his observations of the location in his register in subsequent rounds of testing.

about flaw location.

• Keyman to watch during daily patrolling till it is joggled fish plated.


Various types of rail sections and their specifications have been discussed in this chapter. Flat-footed rails are commonly used on Indian Railways. A rail section may develop different types of defects during its service life. The defect in a rail should immediately be attended to, otherwise it will lead to the failure of the rail. The various types of defects in a rail and the remedial measures to be adopted have been highlighted in this chapter.

Review Questions

1. What are the functions of rails? Name the various types of rails in use. Which one is widely used now? How is the weight of a rail section usually determined?

2. Enumerate the various stresses a rail is subjected to in a track. Describe the nature of contact stresses between the wheel and the rail, distinguishing the cases of new and worn wheels.

3. It is observed that at present tracks, are mostly laid with flat-footed rails. Give reasons for this preference in relation to other types of rail sections.

4. Defects in rails can be divided into the following three ctaegories:

(a) Defective rail steel

(b) Surface defects

(c) Service defects Explain these defects clearly.

5. What is meant by wear of rails? Categorize the types of rail wear and enumerate the methods by which wear in rails can be measured.

6. What factors govern the permissible limit of rail wear?

7. Determine the suitable rail section for a locomotive carrying an axle load of

22.5 t. (Ans: 90 R)



Rail Failure | RAILWAY ENGINEERING | Sleepers intro