No abstract

No abstract

The numerical model is a well-acknowledged tool to evaluate railway pantograph-catenary interaction performance. The current standard restricts most current simulation tools by a cutoff frequency of 20 Hz. This low-frequency range of interest cannot fully describe the current collection quality of pantographcatenary. This paper includes simulation with cutoff frequencies up to 200 Hz to investigate the high-frequency behaviour of pantograph-catenary interaction. The reference model of pantograph-catenary in the benchmark is taken as the analysis object. Firstly, the effect of key simulation parameters of the resulting contact force is investigated. A small element length in the finite element model is proposed to prevent the frequency range of interest being contaminated by the numerical error. The contact stiffness has an opposite effect on the contact force in low and high-frequency ranges. Then the source and the amplification factor of high-frequency components of contact force are investigated. The results show that the half and quarter of the dropper/steady arm interval length presents the primary source of high-frequency components of the contact force. The corresponding wavelength can also be found in the high-order modes of the catenary. Finally, a variable time step procedure is also proposed to capture the contact loss occurring at high frequencies accurately. The comparison of the results between the variable and constant time steps indicates that the traditional constant time step may result in errors when calculating the contact loss duration.

The contact force between the pantograph and the contact wire ensures energy transfer between the two. Too small of a force leads to arching and unstable energy transfer, while too large of a force leads to unnecessary wear on both parts. Thus, obtaining the correct contact force is important for both field measurements and estimates using numerical analysis. The field contact force time series is derived from measurements performed by a selfpropelled diagnostic vehicle containing overhead line recording equipment. The measurements are not sampled at the actual contact surface of the interaction but by force transducers beneath the collector strips. Methods exist for obtaining more realistic measurements by adding inertia and aerodynamic effects to the measurements. The variation in predicting the pantograph-catenary interaction contact force is studied in this paper by evaluating the effect of the force sampling location and the effects of signal processing such as filtering. A numerical model validated by field measurements is used to study these effects. First, this paper shows that the numerical model can reproduce a train passage with high accuracy. Second, this study introduces three different options for contact force predictions from numerical simulations. Third, this paper demonstrates that the standard deviation and the maximum and minimum values of the contact force are sensitive to a low-pass filter. For a specific case, an 80 Hz cut-off frequency is compared to a 20 Hz cut-off frequency, as required by EN 50317:2012; the results show an 11% increase in standard deviation, a 36% increase in the maximum value and a 19% decrease in the minimum value.

An increasing demand for reduced travel times requires the exploitation of the full capacity of existing overhead railway catenary systems. This need has become an issue in Norway, as the majority of existing catenary systems are designed for a maximum speed of 130 km/h. In many regions, plans to reconstruct the railway line do not exist. Therefore, existing catenary sections must be optimized to increase a train's velocity and reduce the total travel time. In this paper, the dynamic response is evaluated in an optimization investigation of an existing soft catenary system. A dynamic investigation that considers finite element models of existing soft railway catenary sections with original tension forces, current tension forces and suggested new tension forces for velocities at and above the design speed is conducted. The dynamic response is quantified by the interpretation of spectral densities and variations in their peak values. Due to more movement at mid-span than at the pole support, the effects from altering the tension forces and increasing the speed can be more accurately described and estimated by considering the dynamic content of the response at mid-span instead of the peak uplift at the pole support. A 23% increase in speed is possible for the system with the best tested new tension force setting, in which only the dynamic response and uplift at the pole support are considered.