Reliable estimates of the small-strain stiffness of railway ballast are essential for modeling train-induced vibration transmission and supporting condition assessment in underground railways. This paper presents a Multichannel Analysis of Surface Waves campaign performed inside a tunnel of Milan Metro Line M1, using a controlled impact source, a setting not yet discussed in the literature. A short, densely sampled receiver array was installed along the track, Rayleigh-wave dispersion was obtained in the frequency–phase-velocity domain using a phase-shift approach, and a 1-D layered shear-wave velocity V s profile was obtained by inversion. A key methodological insight derives from repeating the survey under two track boundary conditions: fastened and unfastened rails. Comparison of dispersion images and inverted profiles shows that the fastened configuration yields higher phase velocities and V s values, consistent with stronger rail–sleeper coupling and rail-guided energy, which can bias interpretation of ballast properties. Conversely, unfastening the rails suppresses these effects and produces lower misfits and V s profiles more representative of the ballast–invert–subgrade system. Three progressively constrained inversion parameterizations were tested to address non-uniqueness and robustness of the obtained profiles. The results confirm that Rayleigh-wave dispersion is primarily controlled by V s , whereas other characteristics remain weakly correlated. Finally, in-situ stiffness trends of ballast are benchmarked against a laboratory dataset on comparable materials, supporting the plausibility of the velocity profiles obtained. Overall, the study demonstrates the feasibility of active MASW in a tunnel environment and delivers an operational workflow to obtain ballast-scale V s profiles, while highlighting the importance of controlling track boundary conditions.
Maugeri et al. (Fri,) studied this question.