A field-validated numerical prediction model for railway-induced ground-borne vibrations

Abstract

The pursuit of optimal land usage has endangered a reduction in the spatial separation between buildings and train passages, resulting in heightened vibration exposure for occupants and structures. Prolonged vibration exposure to structures, caused by dynamic train loads, alters their structural properties, and leads to deleterious effects such as damage, deterioration, and an increased risk of failure together with the degradation of serviceability conditions. However, the implementation of diligent vibration monitoring protocols holds promise for mitigating these adverse effects. In this study, a sensor-based accelerometer device called VIBSEN (Vibration Sensing) was developed to measure train-induced ground vibrations in the field. The dependability of the VIBSEN device was proven through gravity calibration tests and by comparing the results with those obtained from a calibrated vibrometer. Furthermore, a three-dimensional (3D) numerical prediction model of a train-track-ground system was developed using MIDAS GTS NX software. This model incorporated rail beams, concrete sleepers, ballast, sub-ballast layers, and subgrade. The acceleration-time history data obtained from the numerical model aligned well with the data obtained from the VIBSEN device during the experimental study. These findings affirm the dependable accuracy of the numerical prediction model and the efficiency of the VIBSEN device in measuring train-induced ground vibrations.