The present invention relates to a method and apparatus for ultrasonic wheel testing and relates particularly, but not exclusively, to an automatic ultrasonic wheel testing facility for detecting defects in rolling stock wheels.

DISCUSSION OF PRIOR ART

The need to be able to detect defects in rolling stock wheels arose out of the applicant's own experience of numerous rolling stock ore car wheel fractures, some of which resulted in catastrophic failure with subsequent derailment. The applicants ore haulage is based on a unit train concept, with three (3) 3,600 h.p. head end locomotives hauling some 200 ore cars of 100 tonnes capacity each, with axle loads of between 30-32 tonnes. These trains are up to 2 kilometres long and may have a gross tonnage exceeding 24,000 tonnes. Track speed is between 70 kilometres per hour and 80 kilometres per hour. Accordingly, if a single ore car wheel set fails, the results are usually catastrophic causing millions of dollars of damage to the rolling stock and track structure.

The majority of catastrophic wheel failures are caused by surface opening atigue cracks either in the wheel tread or wheel flange areas. The surface opening fatigue cracks usually initiate from gouge marks caused by the retarder shoes on the wheel grippers at the ore car dumping facilities, the cracks propagate, and finally attain critical crack size whereupon catastrophic failure results. Another cause of wheel failure is over-heating, due to either drag braking or prolonged braking, which tends to reverse the compressive hoop stress, from the wheel rim heat treatment manufacturing process, to tensile hoop stress.

Subsequent to several catastrophic wheel failures in the early 1970's, ultrasonic examination of rolling stock ore car wheels was introduced, utilising a Wheelfax Junior MKII ultrasonic search unit. As technology improved, a Krautkramer USM2-NF flaw detector was utilised, and a Wheelfax 0.4 MHz low frequency probe generating surface waves which travelled around the wheel rim circumference on the tread /flange area. If the

wheel integrity was satisfactory, then a backwall echo would be displayed on the USM2-NF cathode ray tube (CRT) , indicating that the surface waves had completed a circuit of the wheel circum erence. Any discontinuities in the wheel, however, would reflect some of the surface waves back to the probe search unit, which would then display electrical signals on the CRT as peaks above the base line in front of the backwall echo.

Unfortunately, ^' ultrasonic surface waves, or Rayleigh waves, have a number of disadvantages. Because the sound waves travel along the component surface, any irregularity on the surface be it dirt, a scratch, a thermal crack or even a finger will reflect some of the sound back to the probe search unit, causing a spurious indication on the CRT screen. In addition, surface sound beam penetration is less than one half wavelength under ideal conditions, and as such the method is only considered suitable for detection of discontinuities at, or very close to, the surface. As the rolling stock ore car wheels sustain considerable surface mechanical damage, attenuation or dispersion of the surface wave sound beams is considerable with the system unable to detect surface opening cracks on the rear of the flange. SUMMARY OF THE INVENTION

The present invention was developed with a view to providing a method and apparatus for ultrasonic testing of rolling stock wheels, which would enable detection of surface opening fatigue crack defects prior to attaining critical crack size, and can therefore minimize catastrophic wheel failure.

According to; one aspect of the present invention, there is provided a method of detecting defects in rolling stock wheels, the method comprising the steps of: positioning a rolling stock wheel over ultrasonic probe means; raising the ultrasonic probe means into rolling contact with the wheel; commencing rotation of the wheel; completing an ultrasonic scan of the wheel with the ultrasonic probe means as the wheel rotates; ceasing rotation of the wheel; and,

retracting the ultrasonic probe means; whereby, in use, any fatigue defects in the wheel can be detected during said ultrasonic scan.

Preferably the method also comprises the steps of: spraying liquid couplant onto the wheel immediately prior to or as it commences to rotate;and, automatically shutting off the water couplant spray when wheel rotation ceases.

In a preferred embodiment, the step of positioning a rolling stock wheel over the ultrasonic probe means involves: rolling the wheel over the ultrasonic probe means on a support rail; raising an idler roller to arrest forward movement of the wheel and to locate it over the ultrasonic probe means; lifting the wheel from the support rail using a drive roller; and, withdrawing the support rail from beneath the wheel.

Preferably said step of raising the ultrasonic probe means into rolling contact with the wheel involves: raising a probe platform, provided with a wheel tread probe and a wheel flange probe, until the wheel tread probe makes contact with the tread of the rolling stock wheel; followed by the step of, moving the wheel flange probe into contact with the flange of the rolling stock wheel.

The step of retracting the ultrasonic probe means preferably comprises: retracting the wheel flange probe away from the wheel flange; and, lowering the probe platform below the level of the support rail.

Preferably the method involves testing both wheels of a wheel set simultaneously, by performing each of the aforementioned steps for both wheels of the wheel set concurrently, and preferably the method of the invention is performed with the wheel set in situ on a rolling stock car and a further step of positioning the next wheel set over the

ultrasonic probes preferably involves indexing the car automatically to the required position.

According to another aspect of the present invention, there may be provided an apparatus for detecting defects in rolling stock wheels, the apparatus comprising: a wheel rotation mechanism for rotating a rolling stock wheel; ultrasonic probe means for performing an ultrasonic scan of the wheel as it is rotated by said rotation mechanism, whereby, in use, defects in the wheel can be detected during said ultrasonic scan.

Preferably said ultrasonic probe means comprises an ultrasonic wheel tread probe and an ultrasonic wheel flange probe, said ultrasonic wheel probes being adapted to be brought in to rolling contact with the wheel prior to performing said ultrasonic scan.

The apparatus preferably further comprises liquid spraying means for continuously spraying liquid couplant onto the wheel during rotation to facilitate acoustic coupling of ultrasonic signals from the ultrasonic wheel probes into the wheel.

Preferably said ultrasonic wheel probes each comprise a rotatable wheel fitted with fluid filled tyres housing one or more ultrasonic transducers, said transducers being mounted to a fixed axle on which said rotatable wheel can rotate. Preferably the wheel tread probe is provided with three ultrasonic transducers, two of the transducers being oriented to transmit a sound beam into the surface of the wheel at 70° with respect to a direction perpendicular to the wheel surface, and a third transducer being oriented to transmit a sound beam into the surface at 0° with respect to the perpendicular direction whereby, in use,., optimum strength of signals reflected from fatigue defects in the wheel can be obtained.

In order to facilitate a better understanding of the method and apparatus according to the invention, a preferred embodiment of an ultrasonic wheel testing facility will now be described, by way of example only, with reference to the accompanying drawings. Although the following description relates to a method

and apparatus for testing the wheels of ore cars, it will be apparent that the invention has wider applications, such as,for example, the testing of wheels on rolling stock owned by public railway authorities.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 illustrates in a plan view a rolling stock wheel set, in position over the ultrasonic testing station;

Figure 2 is a side elevation view along the line 2-2 in Fig. 1, illustrating a rotation mechanism and a tread probe in rolling contact with the wheel;

Figures 3 & 4. are end-elevation views along the line 3-3 in Fig. 1, illustrating the ultrasonic probes in a raised and a partly lowered position respectively;

Figure 5 is a schematic part-section view of a preferred embodiment of an ultrasonic wheel probe; and,

Figure 6 is a block diagram illustrating the signal processing of ultrasonic signals from/to the probes.

DESCRIPTION OF THE PREFERRED EMBODIMENT

In Fig. 1 a rolling stock wheel set for an ore car is illustrated in plan view, the wheel set 10 comprising first and second wheels 12, connected to an axle 14. The wheel set 10 is illustrated positioned over the prototype ultrasonic wheel testing station, (not shown) , and the location of first and second wheel flange probes 16 behind each wheel 12 is illustrated in broken outline. In a prototype ultrasonic wheel testing facility, ultrasonic probe means were mounted in pits 18 of a non-operational KSK under-floor wheel mill. The non-operational

KSK under-floor wheel mill provided an ideal wheel set rotation mechanism, with no capital expenditure being required. This technique of rotating the wheels 12 enables in-situ ultrasonic testing of rolling stock wheels. However, obviously other techniques for rotating the wheels can be employed to similar effect.

The wheels 12 are positioned over the ultrasonic probe means by rolling the wheel set 10 on support rails 20. Forward movement of the wheel set 10 as it locates over the ultrasonic probe

means, is arrested by an idler roller which is raised above the level of the support rail 20. Referring to Fig. 2, the idler roller 22 can be seen in the raised position where it arrests forward movement of the wheel 12 as it is rolled into position. When the wheel 12 is in the correct position, a drive roller 24 lifts the wheel 12 from the support rail 20, and the latter is then withdrawn from beneath the wheel 12. The idler roller 22 and drive roller 24 together comprise a wheel rotation mechanism for rotating the wheel 12 during ultrasonic scanning. The revolving speed of a wheel set on the original under floor-wheel mill, was only 0.33 RPM and therefore modifications were required to enable an increase of the revolution rate. As the milling machine drive was hydraulic, and modification to the hydraulic drive was not feasible, the hydraulic motor was disconnected and a 3-phase electric motor and reduction gearbox were substituted to give a maximum revolution rate of 15 RPM. A PDL variable frequency motor control was wired to the electric motor enabling wheel set rotation speeds from 0 - 15 RPM. This enabled defects in the wheel to be pin-pointed for further examination if necessary. Also illustrated in Fig. 2, is an ultrasonic wheel tread probe 26 of the ultrasonic probe means mounted on a probe platform 28 which may be raised and lowered as required. The ultrasonic wheel tread probe 26 is illustrated in rolling contact with the tread of the wheel 12. When the wheel 12 has commenced rotating by the action of the drive roller 24, the ultrasonic wheel tread probe 26 completes an ultrasonic scan of the wheel tread as the wheel rotates through a complete revolution. A water spraying means 29 is also provided for continuously spraying water couplant onto the wheel during rotation to facilitate acoustic coupling of the ultrasonic signals from the ultrasonic probe into the wheel. This use of a water couplant is preferable because the ultrasonic sound beam is reflected from, and/or greatly attenuated as it propagates through, any air gap it encounters in its path. As an alternative to spraying, the probe platform 28 with probes mounted thereon could be fully immersed in water when the probes make contact with the wheel 12.

However, this latter arrangement is more complicated and expensive than simple water spraying means.

When the ultrasonic probes have completed a scan of the wheel 12, the probe platform 28 is lowered into the pit 18 below the level of the support rail 20. The support rail 20 is moved back into position beneath the wheel 12, and then the drive roller 24 lowers the wheel 12 down onto the support rail. The idler roller 22 is ^' then also lowered to allow the wheel 12 to be rolled away from the ultrasonic testing station. This sequence of operations is performed simultaneously on both wheels 12 of the wheel set 10 so that ultrasonic testing of one wheel set (two wheels)can be completed in the time it takes to complete one revolution of the wheel set.

Referring now to Figs. 3 & 4 which are end elevations along the lines 3-3 in Fig. 1, illustrating the ultrasonic probes in a raised and a partly lowered position respectively (the wheel rotation mechanism and water spraying means have been omitted for clarity) . In Fig. 3 the probe platform 28 is in a raised position with the ultrasonic wheel tread probe in rolling contact with the tread of the wheel 12. The position of the wheel flange probe 30 in rolling contact with the flange of the wheel 12, can also be clearly seen in Figure 3. The probe platform 28 is pneumatically operated from the workshop air supply. The wheel tread probe 26 is mounted on an axle held in a U-shaped cradle 32 which is bolted directly to the probe platform 28. The wheel flange probe 30 is likewise rotatably mounted on an axle held in a U-shaped cradle 34. However, the wheel flange probe is provided with a small independent pneumatic cylinder 36 which is automatically activated when the wheel tread probe 26 makes rolling contact with the tread of the rolling stock wheel 12. In this way any slight differences in wheel set back-to-back displacement can be accommodated by the wheel testing station. At completion of the test, the flange probes 34 withdraw and the platforms 28 retract below the support rail area into the pits 18 of the KSK mill. Water sprays (not shown) to enable coupling of the sound from the ultrasonic wheel probes into the rolling stock wheel are automatically activated when the ultrasonic wheel tread probes

26 touch the rolling stock wheels 12. Leads 38 and 40 shown in Fig. 3 & 4, carry electrical signals to and from the ultrasonic transducers within the wheel flange probe 16 and the wheel tread probe 26 respectively, for processing. Each of the wheel probes employed in the prototype ultrasonic wheel testing station comprises two wheel hubs 42 which support a resilient tyre 44 within which a liquid water/glycol mixture is sealed to provide a liquid interface between the ultrasonic transducers housed within the wheel probes and the rolling stock wheel 12. In this particular embodiment of the wheel testing station, the ultrasonic wheel flange probes differ from the wheel tread probes in that the wheel flange probes contain one 70° transducer, whereas the wheel tread probes contain two 70° transducers and one 0° transducer. In Fig. 5 the internal configuration of a typical wheel tread probe is illustrated showing the position and orientation of the ultrasonic transducers.

Referring to Fig. 5 a preferred embodiment of a wheel tread probe is illustrated schematically in part-section view. The probe comprises one plain and one valved housing half 42 which fit onto a hollow axle 43 either side of an axle block, (not shown) , on which the ultrasonic transducers are mounted. A polyurethane tyre 33 fits over the two housings 42 and is clamped by two tyre clamps 45. The probe assembly is filled with a glycol-water mixture 41 through two valves in the valved housing, one valve venting the air while the fluid is slowly admitted through the other. Air removal is an important part of the assembly process. A glycol-water mixture is chosen to provide constant velocity of sound over the intended operating temperature range.

The axle blocks, (not shown) , are provided with slots to mount the transducers, and hollows to accommodate the transducer signal cables and to lead them out through the hollow axle 43. The wheel tread probe is provided with three ultrasonic transducers, two transducers 46 being oriented to transmit a sound beam into the surface of the wheel at 70° with respect to the perpendicular direction 47 and the third transducer 48 being

oriented to transmit a sound beam into the surface at 0° with respect to the perpendicular direction 47. The 70° angle was carefully chosen following trials and experiments to obtain optimum strength of signals reflected from any fatigue defects in the wheel tread surface. Unlike the surface of a rail which is linear, the surface of the wheel tread 12 is curved and therefore the transmission angle of the sound beam becomes more critical to reliably detect defects.

The majority of catastrophic wheel failures occurred with the fatigue crack initiating in some form of mechanical surface damage and propagating in a radial direction, with the average critical crack size being 17mm. The wheel probes were tested on a test wheel set with simulated defects in the form of 3mm diameter drilled holes and saw cuts in the areas of the wheels where fatigue cracks occurred in service. Although the reflecting surfaces of the drilled holes represented a simulated discontinuity size of less than 1mm, these simulated defects could be detected quite satisfactorily with the wheel probe. Transducers with a frequency of 5 MHz were used in the wheel probe, however, other frequencies can also be utilized. Rr example, a frequency of 2.25 MHz was used successfully during trials.

In Fig. 6 signal processing apparatus for processing the signals from/to the ultrasonic probes is illustrated schematically in block diagram form. The two pairs of ultrasonic probes 16 and 26 are shown in rolling contact with the East wheel and the West wheel respectively of the wheel set 10. In the illustrated testing station the wheel tread probes 26 have three transducers each, and the wheel flange probes 16 have two transducers each (70° & 0°, but only the 70° transducer is used, 0° is used for calibration of 70° transducers) , making eight ultrasonic transducers that are used in all. The eight ultrasonic transducers from the four wheel probes feed into an eight channel Krautkramer Multiscanner 50 which is driven by a Sonatest flaw detector 52. The flaw detector 52 operates one channel of the multiscanner at a time and provides a visual display, (on a CRT in the front panel) of the signals transmitted to each of the

ultrasonic probes together with traces of any reflected signals detected by the probe transducers. The eight channels can be displayed simultaneously or individually as required. Each channel has an illuminated flaw light capability which is also connected to an audible alarm. When, a defect is detected the transducer from which the reflected signal originated is isolated and an ultrasonic scan of the corresponding wheel 12 is repeated with the wheel 12 rotating more slowly in order to pin-point the location of the defect. From the above description of the wheel testing facility it will be evident that the method and apparatus according to the invention lends itself to being automated in order to facilitate continuous testing of a large number of wheel sets.

In this connection, the described ultrasonic wheel testing apparatus will be connected to a Programmable Logic Controller (PLC) for automatically controlling the operation sequence according to the method of the invention. The PLC will also act as an interface to a computer based wagon maintenance system. The PLC will automatically control the indexing of the wheel set, the raising of the wheel set, the retracting of the support rail, the raising of the probe platform, the scanning sequence during rotation of the wheel and the subsequent steps as described above. In addition, the PLC will also monitor the flaw detector for defect detection, and will interupt the control sequence to signal an operator when a defect is detected in order to enable the operator to take appropriate action, such as precise location of the defect detected. Details of the configuration and connection of the PLC have been omitted from this description as these or a similar control system may be readily designed and implemented by any skilled electronics/control systems engineer. A prototype testing facility employed a capstan winch in order to pull the ore cars into the wheel testing station and index them on to the KSK wheel mill. Current testing rate is approximately 12 cars per day, but this will be increased as factors limiting the throughput, such as the capstan indexing system and siding storage capacity at the wheel mill are addressed. It is envisaged that a fully automatic car inspection

station will be built, where complete ore car trains can be drawn through the station.

The described ultrasonic wheel testing facility is capable of detecting surface opening fatigue cracks well before they reach a critical size. Although rolling contact subsurface fatigue defects are relatively uncommon in applicant's rolling stock, due to the favourable assymetrical rail profile grinding and complementary modified wheel tread profiles on rolling stock wheels, it will be apparent that the ultrasonic wheel testing facility is also capable of being able to detect this type of defect.

In the envisaged fully automated testing station, other facilities could be provided in addition to ultrasonic wheel scanning, such as, for example, residual wheel stress measurement, back to back measurement, roller bearing inspection, wheel profile measurement, bogie geometry measurement and general ore car body visual inspection. It is envisaged that the testing station will be connected to the computer based wagon maintenance system presently in use at applicant's rolling stock maintenance department and that the ultrasonic signal responses will be monitored, recognized and recorded by a computer interface system.

From the above description of a prototype wheel testing facility, it will be apparent that the method and apparatus according to this invention provide an improved technique for the detection of defects in rolling stock wheels, which enables the detection of surface opening fatigue cracks well before they reach a critical size so that the wheel set can be withdrawn from service prior to catastrophic failure. It will be apparent to those skilled in the relevant technical fields, that numerous variations and modifications may be made to the described prototype testing facility, without departing from the basic inventive concepts. All such variations and modifications are to be considered within the scope of the invention, the nature of which is to be determined from the foregoing description and the appended claims.