GNGTS 2017 - 36° Convegno Nazionale

GNGTS 2017 S essione 3.3 699 near as possible to the railway track, in order to evaluate the soil attenuation and the amount of vibrations reaching the rock mass. The accelerometers were connected to a Kinemetrics K2 datalogger, while the velocimeters were connected to ������� ��� ��� ������� ����������� Trimble REF TEK 130S-01 ����������� dataloggers set to record with a sampling frequency of 250 Hz, with K2 in trigger mode and REF TEK in continuous mode. Data processing. Because of limited storage capacity, the accelerometric records are discontinuous and do not include all the trains transited during the experimentation period. Anyway, the collected data are adequate to perform preliminary analyses and to test the methodology of record disaggregation in different frequency ranges. The accelerometric records were filtered from 4 to 60 Hz, and then integrated to obtain velocity records in order to be comparable with those acquired by Lennartz velocimeters. Fast Fourier Transform (FFT) and spectrograms computed for the transit of different train typologies (i.e. freight trains, passenger trains) show a peak around 50 Hz, with a FFT amplitude that is similar for the accelerometers 5-6-12 and about an order of magnitude smaller for sensor 8. Mean ambient noise levels were obtained in three frequency ranges (low: 0.1-30 Hz; medium: 30-50 Hz; high: 50-60 Hz) to measure the amount of vibration that reaches the rock mass in each frequency range. The preliminary results evidence that the rock mass vibrational behaviour is mainly influenced in medium and high frequency ranges. In order to obtain a better estimation, further analyses will include more frequency ranges. Conclusions. The two here-presented test sites can be considered as natural scale laboratories for investigating the effects of induced vibrations on fractured rock masses. In Acuto test site it was possible to observe the vibrational response of a rock block and its independent mobility respect to the behind rock wall, evidenced by SSR analysis and spectrogram results as well as by the three-dimensional micrometric displacement obtained during shaking tests. In Terni test site it was possible to record numerous events of induced vibrations, which will be processed in order to establish the influence of environmental parameters as well as of train typology on variations of the mean ambient noise trend. Further data analyses will be performed to understand whether the dynamic inputs of the Fig. 2 - Terni-Giuncano railway line test site. Positioning and ID number of the accelerometers (green circles), velocimeters (yellow circles) and of the K2 datalogger (red circle) are shown. REF TEK dataloggers were placed in an impermeable box (bottom right).