PSI - Issue 17

João Morais et al. / Procedia Structural Integrity 17 (2019) 448–455 João Morais et al./ Structural Integrity Procedia 00 (2019) 000 – 000

452

5

a)

b)

c)

d)

Fig. 3. Installation of the instrumentation: a) ±500 g accelerometer without its protective cover; b) connection boxes installed near a wheels shaft; c) welding of a support for one displacement transducer; d) metal cutting operations for wiring and ±4 g accelerometer installation.

4. Prototype system calibration and assessment

System calibration and validation tests were performed. To this end, a railway section adjacent to the maintenance yard was selected and prepared for this purpose. Several additional systems and instrumentation were installed to monitor the structural response of the selected railway track and serve as reference. It is known that measuring railway deformations is a difficult task to achieve, even with on-site instrumentation. The simplest solution for measuring railway deformation would be the use of displacement transducers. However, this solution requires the establishment of a reference point (static, outside the interference zone), to be used by the displacement transducers. Since the area surrounding the rail also suffers deformation, this reference point must be set at a distance such that it is not affected by the load imposed on the rails. Under these conditions, the reference point would have to be established at a minimum distance of two to three meters, which would create other constraints and make this solution unfeasible. So it was decided to implement a vision type non-contact monitoring system, based on a video camera and a set of LED active targets. The railway track was also instrumented with a redundant system to measure the applied deformations, based on high-sensitivity accelerometers, with the ability to measure acceleration in static conditions (Fig. 4).

W

E

Distance between sleepers (cm)

Sleepers

LEDs

Accelerometers

Fig. 4. Diagram representing the instrumentation positions in the railway track used for the calibration tests.

The vision type non-contact monitoring system, based on computational vision techniques, is composed of an industrial-type video camera with 1920 × 1080 pixel resolution. In order to minimize the effects of ambient lighting,