Table of Contents ECNDT '98 Session: Railway Copenhagen 26 - 29 May 1998

Ultrasonic Evaluation of Stresses in the Rims of Railroad Wheels E. Schneider *, R. Herzer Fraunhofer Institut Zerstörungsfreie Prüfverfahren (IZFP) Saarbrücken, Germany * Corresponding Author: Eckhardt Schneider Head of Department Process integrated Nondestructive Testing at Fraunhofer Institut Zerstörungsfreie Prüfverfahren IZFP, University Bldg 37, D-66123 Saarbrücken Phone: +49 681 302 3840 Fax: +49 681 39580 Email: schneider@izfp.fhg.de Homepage: http://www.izfp.fhg.de/english/ NDTnet: Q Net Stand

Hear Mr. Schneider live in Copenhagen . You may study the viewgraphs while listening. (This is a 20 minutes speech as a streaming Real Audio file. Download the Player) The transcript of the speech coming soon. Session pictures: - -

Summary

The stress states in railroad wheels are strongly influenced by the heating and cooling process during and after the brakings using brake shoes. Different ultrasonic set-ups to evaluate the stress states in the rims of monoblock railroad wheels are already in use. Although developed by different institutions, the technique used by the set-ups is the same. By measurements of the times-of-flight of an ultrasonic shear wave polarized along the radial and along the circumferential direction, the difference of the principal stresses acting in the circumferential and in the radial directions (_circum. - _radial ) is calculated. The application of this ultrasonic technique presumes that there is no other reason for the times-of-flight differences of the shear wave polarized along the two directions than the stress state. But in numerous older wheels, which are still in use, the manufacturing process caused a texture in the rim. The texture also influences the times-of-flight of the shear wave polarized along the two principal axis. Hence, the stresses evaluated for textured wheels are not reliable.

Texture is often going along with a significant direction dependence of the grain dimensions resulting in direction dependent ultrasonic attenuation. The direction dependences of the ultrasonic attenuation and of the shear wave time-of-flight have been measured using more than 60 wheels of the older types. It is found that both quantities change significantly if there is a texture in the wheel and small changes are observed for wheels without texture or with a slightly developed texture. A threshold value is defined to separate wheels with a strong texture from those with no or with a slightly developed texture. The stress analysis for wheels with a strongly developed texture is only possible if the texture is homogeneous along the circumference of each wheel. The characterization of the stress state in wheels with a slightly developed texture is possible as first results show.

Situation

Distributions of the tangential (circumferential) stresses (tan in cross sections of rims of used monoblock wheels are given in Fig.1. The left part shows the stress profiles in a wheel, heavily braked in a braking test stand while the profiles of a wheel used in traffic is shown in the right part of the figure. The highest tensile stresses in railroad wheels have been found in the edge area of the rolling surface and at the outer side of the rim. The stresses were evaluated using a totaly destructive cutting and sectioning technique. The points give the positions and the results of the strain gauge measurements; the lines are numerically evaluated. The result is regarded as representative with respect to the stress distribution; the absolute values are of course strongly depending on the braking conditions. With respect to the fracture mechanical evaluation of the cracks in that area, the evaluation of the stress states in the surface near layers of the running surface and of the outer side is most desirable. But there is no suited technique available.

Fig.1: Circumferential stress [MPa] and stress profiles in a wheel, heavily braked in a braking test stand (left part) and in a wheel used in traffic (right part). European Railroad Research Institute: Document ORE B169/RP2, Utrecht (1989).

Ultrasonic techniques enable a fast and nondestructive evaluation of stress states in the rims of railroad wheels. The applied approach is based on the fact that changes of the stresses in the surface near areas yield also changes of the stress state in the bulk of the rim. Hence, the velocities of ultrasonic waves propagating the bulk of the rim are influenced by the stress state and its change. The ultrasonic beam is also illustrated in Fig. 1, showing the uppermost and deepest measuring position. The application of a shear wave, propagating the width of the rim, vibrating along the height (radial direction of the wheel) and then vibrating along the circumferential (tangential) direction results in the evaluation equation:

_tan - _rad = K (t_rad - t_tan.) / t_tan.

_tan and _rad are the principal stresses along the tangential and radial direction, respectively. t is the time-of-flight of the shear wave vibrating along the direction indicated as index. K is a material dependend constant, evaluated in the laboratory using representative material samples. The technique is described in more detail by Schneider et al (1994).

As to be seen in Fig.1, the sound propagates the whole width of the rim. The result of the investigation is a mean value of the principal stress differences along the ultrasonic path. Depending on the material of the different types of wheels, a maximum value of acceptable stress difference has been defined in order to assure the safety of the wheels during their service life between the periodical inspections. Besides the material dependent quantities and the results of fracture mechanical investigations, the characteristics of the detected cracks (number, length, orientation) is taken into account, as explained by Rode (1994). The German Railroad (DB AG) specified maximum allowable values of _tan - _rad for different types of wheels. A nondestructive technique to evaluate the crack depths in wheels is developed by Salzburger and coworkers (1996).

As already mentioned earlier, there are different set-ups available to evaluate the stress states in rims using the same ultrasonic approach. The DEBRO 30 and DEBBIE set-ups developed by Deputat (1989 and1995), the set-up developed by the French Railroad SNCF presented by Limal (1995) and the UER set-up developed by Herzer and Schneider (1994) are the ones known to the authors.

Figure 2: Change of stress state with the distance from the running surface of a wheel heavily braked immediately before the stress analysis (upper part) and of a wheel heavily braked long before the stress analysis (lower part).

One system, the UER-system evaluates the principal stress differences _tan - _rad continuously along a radial trace. That allows the localization of the extrem stress values in terms of the distance from the rolling surface. It is obvious that the stress profile depends on the braking conditions as well as on the load history of the wheel. Fig. 2 displays in its upper part a result found in a wheel which was heavily braked just recently before the stress analysis was performed. The stress maximum is right in the surface near zone. The lower part of Fig. 2 shows the stress profile of a heavily braked wheel which was in service for some time after the braking. The load and the influence of lower braking temperatures caused a decrease of the surface near tensile stress.

The major part of the monoblock wheels inspected so far are forged and heat treated. This procedure has, among others, the advantage that there is no significant texture. Investigations showed that there is a slight texture in some wheels, but the texture caused error in the result of the stress analysis is less than about ( 15 MPa.

But in the recent past, more and more older wheels and wheels made from other materials have been measured and a large number of wheels could not be inspected by the UER system. Investigations show that some of these wheels have coarse grains or large segregations or other microstructural anomalies causing a very high ultrasonic attenuation. Some other wheels are found to have a strong texture which changes in strength at different positions of the same wheel. Other wheels have a texture which is uniformely developed along the circumference of the wheel. Other wheels again have a slight texture, homogeneously developed around the circumference.

Approach for the automated testing of inhomogeneous and textured wheels

The UER measuring routines and the software has been improved in order to enable the testing of at least some of the textured wheels also. The UER system measures the ultrasonic signals amplitude as function of the distance from the running surface. In case the amplitude profile is found to be like the solid line in the Fig. 3, the next step of the measuring routine is made. The evaluation of the stress state is possible in the range from about 8 mm till about 55 mm. If the amplitudes are smaller, as shown as dashed line in the same figure, the stress analysis is only made in the depth area in which the amplitudes are well above a given threshold. In case the amplitudes are found to be smaller than the threshold as shown in Fig. 4, the UER system shows the information Microstructural Anomaly in the display of the system.

In order to recognize textured wheels, the amplitude and the time-of-flight of the shear wave are measured during the automated change of the shear wave polarization direction from the radial (0°) into the circumferential (90°) direction.

Fig. 3: Change of ultrasonic signals amplitude with the distance from the running surface measured using a wheel of a new type (solid line) and of an older (dashed line) type.	Fig. 4: Change of the ultrasonic signals amplitude with the distance from the running surface measured using a wheel of an older type.

In Fig. 5 the variations of the ultrasonic signals amplitude with the change of the shear wave polarization direction are displayed. The significant change as shown in the left part of Fig. 5 is found using wheels with strongly developed textures. The result as displayed in the right part of the figure is a typical result for wheels without or with a slightly developed texture.

Fig. 5: Ultrasonic amplitude versus the change of shear wave polarization from the radial (0°) into the circumferential (90°) direction as measured using a wheel with a strongly developed texture (left part) and using a wheel with no or a slightly developed texture (right part).

Based on results of 60 wheels a threshold value is defined to separate wheels with a strong texture from those with no or with a slightly developed texture. The UER system indicates a wheel with a strong texture by the information Textured Wheel in the display. The stress analysis on wheels with a strongly developed texture is only possible if the texture is homogeneous along the circumference of each wheel and if the texture influence on the measuring quantities is within only small variations for wheels of the same type or for wheels of the same manufacturing periode.

The characterization of the stress state in wheels with a slightly developed texture is possible as first results show. The texture influence in one type of wheels is found to be always in the same range of magnitude. Hence, the texture influence can be taken into account by a correction term and the stress analysis yield as result Tensile Stress Smaller Than xxx MPa which again is shown in the display of the UER system.

It will certainly not be possible to separate the texture influence in each individual wheel and to evaluate the stress state with the accuracy of about ( 20 MPa as it is possible using wheels without texture and without other microstructural anomalies.

The additional measurements of the ultrasonic amplitude and the time-of-flight as function of the shear wave polarization direction is automatically started by the UER system if the usual evaluation procedure results in stress values which are not within certain predetermined limits. The additional measurements prolong the inspection time for one wheel from about 1.5 to about 2.25 minutes.

Fig.6: The base unit and the manipulator of the portable UER-T set-up for the automated evaluation of stress states of railroad wheels.

Two kinds of UER set-ups are available. One is designed for the daily use in workshop environments. 20 systems of that type are used in workshops of the Austrian, German and Swiss Railroads since more than 5 years already. The other type is a new development. It is portable to be used for inspections of endangered wheels on the track. Both types of UER systems are identical in terms of measuring technique, data evaluation and documentation. And both types use electromagnetic ultrasonic transducers (EMAT´s) with the advantage that no coupling medium is needed and the manipulation of the ultrasonic sensor along the measuring trace is significantly facilitated ( Wilbrand et al 1989). Fig.6 shows the portable UER system during the inspection of a wheel.
Some technical characteristics of the portable UER-T system are
Hardware

AT- compatible PC 10.4'' TFT color monitor Serial interface for data transfer Parallel printer interface Rechargeable batteries Stand for scanner and reference material for functional check.
Software
Data aquisition and evaluation software Adjustable testing and evaluation parameters Variable material and wheel set parameters Comparision of the results with pre-set limits for tolerable stress states Documentation of all results Menue in German, English, French or Italian

References