Conference Agenda

Elastomeric bearings are essential bridge components designed to transmit vertical loads while accommodating rotations and displacements [Sétra, 2007]. Their design, governed by the NF EN 1337 standard [SNAC, 2021], relies on dimensioning methodologies (Sétra Guide) using a conventional shear modulus fixed at 0.9 MPa in the French regulatory context [Sétra, 2006], in order to ensure durability by limiting distortions.

However, structural monitoring reveals that elastomeric bearings suffer from a lack of assessment methods of the Non-Destructive Testing (NDT) type. Indeed, there are currently no tools available to characterize the condition of a bearing other than visual inspection, the characteristics of which are defined by current technical guideline (REF ITSEO AA). Visual inspections applied to elastomeric bearings present a severe, highly conservative analysis, which tends to underestimate the real residual capacities and, consequently, generate significant maintenance costs.

The instrumentation method consists of measuring the vibratory behavior of the bearings under ambient vibrations. On-site measurement of a frequency spectrum characterizes the real actual behavior of the bearings. This spectrum is analyzed to identify the cut-off frequency of the bearing and to deduce, via inverse analysis, a realistic and robust estimate of the shear modulus value. This parameter is necessary for all regulatory verifications required by the codes in force.

The APAVE case study, conducted without impacting the operation of the structure, highlighted significant wear of the bearings with quantitative damage data. As the bearing no longer fulfills its role as a flexible isolator, this results in the transmission of unanticipated forces to the structure, which may explain the appearance of structural pathologies [Fragnet, 2021].

Given this observation, the APAVE audit concludes that visual inspection may be deemed insufficient to qualify the structural condition of elastomeric bearings. The integration of in-situ dynamic measurements from the moment of the structure's commissioning - aiming to establish a baseline assessment of the bearings - followed by regular verifications, could offer a clear and up-to-date view of their aging. This detailed monitoring would allow not only a better understanding of the evolution of the bearings' condition, but also the determination of their optimal service life and the opportune time for replacement. Thus, such a proactive approach would contribute to preserving the integrity of the structure by anticipating maintenance needs and ensuring effective management of its longevity. The research project may be supplemented by a technical regulatory analysis regarding the results obtained.

Switches and crossings (S\&C, turnouts) are one of the most maintenance-intensive components in the railway system, making condition monitoring and preventive maintenance an attractive option to minimize costs and traffic disruptions. One condition monitoring solution developed for fixed crossings is an accelerometer mounted on the sleeper next to the crossing transition. The ballast condition can then be observed via track displacements obtained via double integration of accelerations, and the crossing running surface condition can be assessed by estimating the dynamic wheel/rail contact forces from accelerations. An advanced way of doing the latter is to solve a linear inverse force estimation problem using Green's function kernel methods (GFKM). By reconstructing the structural displacements, the wheel/rail contact force is estimated by multiplying the displacements via associated impulse response functions (Green's functions) of the track derived from a simulation model calibrated to the turnout of interest.
The objective of the present study is to evaluate the feasibility of applying this method also in swing nose (movable point) crossings for high speed applications, where the nose can be moved to reduce or eliminate the crossing discontinuity (the crossing gap) which would otherwise be present. These crossings constitute a particular challenge as the swing nose is kept in place by contact surfaces and point machines that introduce non-linearities into the track dynamics. Additional objectives include the evaluation of optimal sensor configuration (number, rail and sleeper mountings, etc.) and quantifying the relative movement of the swing nose with respect to the wing rail casting.
To gather the requisite data, a swing nose crossing on a European high-speed rail line was instrumented with 10 state-of-the-art accelerometers and a displacement sensor for verification of the integrated acceleration data. Furthermore, a detailed 3D-scan of the crossing geometry was obtained to get an accurate description of the running surface. With knowledge of the turnout design and the scanned running surface, a Simpack multibody simulation model with a flexible track structure was built, whereafter the model's track properties were calibrated to the measured track responses. The impulse response functions required for the GFKM were then calculated from the calibrated track model, and the inverse estimation of wheel/rail contact forces was performed from measured track responses. The estimated wheel/rail contact forces were verified with those obtained via simulations of dynamic vehicle/track interaction in the calibrated model.
From this study, it can be concluded that there is promise in applying the GFKM also for condition monitoring of swing nose crossings in addition to fixed crossings. The track dynamics are non-linear, but the swing nose is pressed down under the wheel load, making the track structure fairly linear at the loading points of interest. However, it should be noted that the method is sensitive to the data used and requires track model calibration or adaption for the specific crossing to correctly determine the wheel/rail crossing forces.