PSI - Issue 17

Paulo Morais et al. / Procedia Structural Integrity 17 (2019) 419–426 Paulo Morais et al. / Structural Integrity Procedia 00 (2019) 000 – 000

421

3

Regarding the identification of railway defects that affect the vehicle-railway dynamic interaction, this is accomplished by measuring the accelerations transmitted to the vehicle with accelerometers installed on each wheelset and on both cabins. The developed system is also capable of measuring railway physical parameters that are usually not evaluated by conventional auscultation vehicles. In particular the acceleration data can be used to perform a modal analysis and identify the frequencies related with the test vehicle and the railway. It must be pointed out that the system is intended to perform under normal vehicle operation conditions and acquire and process data in real-time. Under these conditions the system is expected to have behavior changes in all of its major components, due to mass variations or degradation of parts in the test vehicle. To assess the quality of the measurements made by the prototype system installed in the vehicle, it was essential to develop a monitoring campaign on a stretch of railway under current operation. The selected stretch is located between Ermidas do Sado and Poceirão, on the South Line. To further verify these results, a specific railway site was equipped with on-site instrumentation. The selected site was the north abutment of the bridge over the Ribeira de São Martinho, on the PK 63+472 (Fig. 1), where a previous LNEC study had been carried out (Paixão et al., 2018), using different types of under sleeper pads (USP) in order to reduce stiffness transitions. The measurement of the railway response where these elements were installed was particularly relevant for the present project since those components introduced significant vertical stiffness variations, which is precisely the parameter to be measured with the developed prototype. The planning of these tests was performed with Infraestruturas de Portugal (IP). 3. Prototype system demonstration on a railway line under current operation

Longitudinal section

Spans = 28.4 m

20 m

3 m

1 2 3

Sleepers

USP Type II (50 sleepers)

Rail

USP Type I (34 sleepers)

2

1

6

3

4

5

7

Backfill height ~ 7m Deck width ~ 13m

UGM 0/31.5mm

CBM BC=5%

Embankment

Bridge deck

Foundation level

Abutment

Piles

- Ballast layer e ≥ 30 cm under the sleeper 1

2

- Sub-ballast layer e ≥ 30 cm

3

- Capping layer e ≥ 20 cm

-Accelerometer on the sleeper

Backwall

USP Type II (50 sleepers)

USP Type I (34 sleepers)

1

6

2

3

4

5

7 8

Joint with deck

CBM

Lateral limit of ballast layer

Ballast wall

Embankment

0/31.5 mm UGM

Track plan view

Side walk

Lateral limit of track plataform

Wing wall

Pile cap

Fig. 1. Site monitored at PK 63+472: diagram showing the longitudinal profile and the on-site instrumentation positions.

The selected railway section was instrumented with eight piezoelectric accelerometers (± 50 g; 100 mV/g), placed over the sleepers, in pre-defined locations, taking into account the different levels of vertical stiffness that would be