USL                     Ultrasonic Sciences Ltd

Up

 

 

 

 

 

Ultrasonic inspection systems for rail wheels and axles

USL manufactures ultrasonic inspection systems for both railway wheels and axles.  There are stringent requirements set by international and national rail authorities for inspection both during manufacture and in-service. USL has provided systems for both types of test and both stages:

  • Inspection of rough machined axles during manufacture in a production cell

  • Testing of both forged and cast wheels in automated manufacturing plant

  • Inspection of hollow axles in-situ on the train.

  • Wheeltread testing in situ on the train

 

Ultrasonic inspection of axles at the manufacturing plant

Normally axles are inspected after rough machining, both radially and "end to end" using immersion methods. The radial test is designed to detect individual defects or clusters of defects which have been elongated during the forging process. The "end to end" test checks the attenuation properties of the axle material, which can give useful information about the structural condition. In addition, it is a quick check for gross centre line defects - if these are found there is no point in progressing to the radial test, which takes much longer.

For the radial test, a group of immersion probes are scanned down the length of the axle, usually using three probes - see below. One of these is a normal incidence probe, the others generate shear waves clockwise and anti-clockwise around the axle. This ensures detection of defects with differing orientations.

 

These probes are mounted on a manipulator which provides X, Z and tip motion. As the scanning head moves along the axle, the Z axis height is adjusted to maintain a fixed water path and the tip axis is adjusted to generate a beam at right angles to the axis of the axle at the blend radii - for example at the sides of the wheelseat. The position of the probe head is controlled by the computer according to the programmed axle shape.

5 channel system used for solid axle inspection after rough machining. The picture shows the loading gantry poised above the immersion tank waiting to remove the axle after inspection.

For automated systems which have to deal with a range of axle types and sizes, we have designed systems with rollers having adjustable separation - see below.

 

 

The rollers are adjusted to the correct position before axles are loaded by a gantry. The ultrasonic computer communicates with the controlling computer to download data about the incoming axle. The inspection can be a completely automated sequence in a manufacturing cell, such as the one shown below. the picture on the left shows a view along the line with incoming axles at the far side of the immersion tank. The loading / unloading gantry and gripper mechanism is shown in the picture on the right.

 

 

 

Automated systems for hollow rail axle inspection

USL first manufactured systems for this application in 1987, when the use of hollow axles was in its infancy. Now an increasing number of trains, especially for high speed railways, are designed with hollow axles. These must be inspected regularly to detect any cracks which may have formed by fatigue during service. These must be detected well before they reach a critical size, when a catastrophic failure can occur.

The defect size to be detected and the inspection frequency is calculated according to crack growth theory, of course with substantial safety margins included. The acceptance standard MUST be specified by the train and axle designer but normally the acceptance criteria is to detect defects 0.5 to 2mm deep, with an inspection interval of several hundred thousand kilometres.

The inspection is usually carried out during a routine scheduled maintenance when there is a very limited amount of time available. The inspection therefore needs to be done without removing the wheelset from the train and in the shortest possible time.

For this, USL developed a system which tests the complete axle from one end only. The minimum number of components are removed from the axle, the inspection unit is coupled to the end of the axle and the test proceeds. The usual inspection time is 10 - 20 minutes, depending on the defect size to be detected and therefore the scan index -ie the axial movement for each revolution of the inspection probes. During this time two shear wave probes rotate continuously as they move down the bore, generating sound beams in both forward and backward directions. This ensures that the full volume of the axle is tested, with no areas being "shadowed" by physical reflectors in the axle.

The inspection parameters, including gain, gate position and activity, thresholds etc are continuously switched according to the programmed axle shape so that the settings are exactly matched to the sound beam distance, the probe orientation and the acceptance standard.

At the end of the test, a "C scan" image is produced showing any significant reflectors - if any reflectors are shown to be above the reference level, the wheelset must be removed from the train for further investigation. A typical C scan display from a "calibration axle" is shown below - this axle has tangential machined defects on the external surface. Defects which would otherwise be hidden underneath the wheels, bearings or brake discs are readily detected.

The USL software includes features for editing the scan settings according to the profile of the axle - see the example below - double click to enlarge.

 

profile edit.jpg (106798 bytes)

 

Defects are automatically analysed and numbered at the end of the scan - all the defect positions can be printed on a test report which includes all relevant details to identify the train, vehicle, wheelset and axle number.

Examples are shown below of some of these systems in use.

 

 

pendolinounder test 5.jpg (232089 bytes)

 

 

Systems for inspection of forged and cast wheels during manufacture.

Rail wheels are manufactured by both casting and forging processes. The basic inspection techniques are similar for both, except that the expected defect position, orientation and type may be different in each case. In general new wheels are tested using immersion or semi-immersion methods, using multiple probes directing sound beams at the wheeltread, or scanning a single probe or probes over the wheeltread area. Some specifications also require inspection of the wheel boss.

International specifications dictate the position and size of artificial defects in the wheels. "Calibration wheels" with these machined defects must be used to setup the system and for regular checks on the effectiveness of the testing system.

The picture on the left shows two multi channel racks systems used for wheel inspection in India. These incorporate 6 or 10 ultrasonic channels with completely independent ultrasonic and gate setups, dependent on the probe position. They are used in conjunction with an automated wheel handling system. The wheel rolls into the testing position and is lowered onto rollers in an immersion tank with the tread section immersed. The wheel rotates and the inspection is completed over 2 or 3 revolutions, after which it is raised onto an outlet conveyor which directs it back onto the production line or into a quarantine section if there are any defect indications.

The ultrasonic computer is fully interfaced with the PLC controller of the handling system.

Other systems have been installed by USL where wheels are inspected ir a horizontal position at a suitable point in the manufacturing sequence. This requires an irrigated probe arrangement, rather than partial immersion of the wheel - an example is shown below.

 

 

 

 

 

 

 

 

The left picture shows part of a machining cell with a wheel mounted on a roller system used for both marking and ultrasonic testing. On the right is shown the probe assembly advancing towards the wheeltread. The radial probe has vertical and angular motion to scan over the tread surface - the couplant flow is turned off in this picture. An axial probe is also housed in the assembly which scans the rim section beneath the wheel.

 

 

 

Special purpose systems - ultrasonic testing of wheels on an underfloor wheel lathe

USL has manufactured a system which is incorporated into an underfloor wheel lathe - this allowed the wheels to be inspected without removing them from the train and inspecting all 4 wheels on a bogie at the same time. The depth of any defects was recorded and this information was used to determine the amount of material to be machined off the tread. (Normally only enough material is removed to bring worn wheels back to the correct profile, but if deeper defects were found the depth of cut could be modified.)

The lathe toolpost was used to scan an irrigated probe over the tread surface as the wheel was rotated by the roller system. Normally wheels are machined with no coolant or lubricant, but for ultrasonic testing a couplant recirculation system was installed. The ultrasonic system was integrated with a Siemens PLC controller for the wheel lathe.

Note that an installation of this type is not possible without complete co-operation from the wheel lathe manufacturer. Preferably the installation should be done on a new wheel lathe during manufacture.

The installation is shown below.

 

On the left a calibration axle is positioned on the wheel lathe, with a Eurostar train waiting to be moved into position. On the right an irrigated probe unit is shown mounted on the toolpost below the wheeltread.