PSI - Issue 4

Uwe Oßberger et al. / Procedia Structural Integrity 4 (2017) 106–114 Author name / Structural Integrity Procedia 00 (2017) 000 – 000

108

3

The prediction of the damage evolution based on visual inspection is also hindered by the fact that a variety of materials is used for the crossing noses. In a recent publication Eck et al. (2013) investigated and compared the material properties 5 different steel grades used in the fixed crossing of a turnout and proposed two previously unused tool steel grades as new fixed crossing material; Fig. 1 shows a crossing where one of 2 new tool steels has been implemented as fixed crossing material. Sophisticated finite element models such as the ones presented by Pletz et al. (2012) and Xin et al (2016) are able to predict local strain, contact pressure and damage for a specific crossing (geometry and material) and load case (consisting of axle load, wheel geometry and speed). These models have been successfully used to compare different material concepts and study the effects of plastic deformation on the material loading, see Pletz et al. (2013). However, these models have two drawbacks: On the one hand they lack the possibility to account for the changes of the actual load condition that consists of the number of train transitions, different wheel profiles, different axle loads and train speeds. On the other hand the computational effort for these models is too high to be implemented in a real time on-site condition monitoring system. In order to derive a model based condition modelling and prediction setup for fixed crossings it is required to combine the knowledge obtained from the finite element models (forward modelling) with signal based condition monitoring (inverse modelling). This paper presents  the first steps, i.e. sensor instrumentation and signal processing, to establish way side condition monitoring of the new crossing nose and  quantitative measurements of the geometry changes of the new fixed crossing material due to wear and plastic deformation in service. As already mentioned in the introduction, a fixed crossing with a new tool steel nose has been manufactured by voestalpine VAE GmbH. Fig. 2 shows pictures of the corresponding crossing right after strain gauges had been installed on the bottom of the crossing nose – measuring bottom bending strain (longitudinal direction). Note that the picture in Fig. 2a shows the crossing turned upside down and steel covers were installed to protect the strain gauges and the signal cables. The strain gauge placement on the bottom of the fixed crossing of the turnout ensured a protection of the strain gauges with respect to possible ballistic impacts by ballast. Fig. 2b shows the instrumented turnout shortly after installation in track at a test site with mixed traffic in the Austrian railway network. 2. Strain measurement and signal processing

Fig. 2. (a) Picture of the instrumentation of the turnout (turned upside down) with strain gauges at voestalpine VAE GmbH (b) Picture of the instrumented turnout shortly after installation in track at the test site, the strain gauge placement on the bottom of the fixed crossing of turnout ensured a protection of the strain gauges with respect to ballistic impacts by ballast The initial strain gauge measurements performed during the first months after the installation in track have been used by Ossberger et al. (2013) to verify the FE-based predictions of load, contact pressures and local plastic deformation in combination with a load calibration of the instrumented crossing. Since the installation in track in 2012 strain gauge signals of single passing trains were recorded in regular intervals to establish a database to develop signal based condition monitoring. Additionally, it can be assumed that ballast deterioration over the past years had an influence on the strain measurements. Therefore, direct dynamic load measurements (as shown by Ossberger et. al., 2013) would demand for repetitive load calibrations.