No abstract

An extensive full-scale field test to investigate the influence of under sleeper pads (USPs) on track quality (with respect to magnitudes of vertical track irregularities) and track dynamics has been conducted on the Schweizerische Bundesbahnen test site at Kiesen in Switzerland. Static track geometry data and dynamic track responses in terms of sleeper strain and accelerations of rail and sleeper during train pass-by have been measured and analysed. It is observed that the use of USPs generates higher rail and sleeper accelerations but lower sleeper strains due to bending. The degradation of track geometry appears to slow down when USPs are used.

The vertical dynamic interaction between a railway vehicle and a slab track is simulated in the time domain using an extended state-space vector approach in combination with a complex-valued modal superposition technique for the linear, time-invariant and two-dimensional track model. Wheel-rail contact forces, bending moments in the concrete panel and load distributions on the supporting foundation are evaluated. Two generic slab track models including one or two layers of concrete slabs are presented. The upper layer containing the discrete slab panels is described by decoupled beams of finite length, while the lower layer is a continuous beam. Both the rail and concrete layers are modelled using Rayleigh-Timoshenko beam theory. Rail receptances for the two slab track models are compared with the receptance of a traditional ballasted track. The described procedure is demonstrated by two application examples involving: (i) the periodic response due to the rail seat passing frequency as influenced by the vehicle speed and a foundation stiffness gradient and (ii) the transient response due to a local rail irregularity (dipped welded joint). ARTICLE HISTORY

Based on the track geometry car recordings performed from 1999 to 2016 on a section of the Swedish heavy haul line Malmbanan, the vertical track geometry degradation is analysed for wavelengths in the interval 1–25 m. The upper layer of the subgrade on parts of the rail section is peat (depths of up to 2 m), while it is moraine on others leading to a significant longitudinal variation in substructure stiffness. The degradation rates of irregularities in the longitudinal level and the influence of track maintenance (tamping) on the track geometry are studied. In parallel, a method for continuous measurement of track vertical stiffness along the line, allowing for the detection of track sections with poor support conditions, is described and demonstrated. Synchronised measurements of the longitudinal level and the track vertical stiffness are evaluated to determine whether there is a correlation between a high stiffness gradient due to variations in substructure stiffness and a high growth rate of local track geometry irregularities. It is shown that recurrent severe local track geometry irregularities often occur on track sections where there is a combination of a low magnitude and a high gradient in the substructure stiffness. In such cases, tamping may not be a cost-efficient long-term solution to the problem. Instead, upgrading of ballast and subgrade layers should be considered as an option. It is concluded that measurement of track vertical stiffness is an efficient method for the maintenance planning of a more robust railway track, which also minimises the life cycle cost and environmental footprint.

The bond between strands and concrete is of importance for prestressed concrete. The research presented in the current paper aims at a better understanding of the bond mechanism, and of how different detailings of the strand interface affect the behaviour. A bond model for three-wire strands was established and calibrated by use of pullthrough tests. The results from finite element (FE) analyses with the bond model and the tests were used in parallel. It was found that adhesion, friction and the ability to develop normal stresses determine the bond response of the strand; consequently, they were used as input parameters in the bond model. How different detailing of the strand surface affects these parameters, and the influence on the bond mechanism, are shown. For example, adhesion has the strongest influence on the initial bond response in the cases of smooth and indented strands. Regarding indented strands, the maximum bond capacity is determined by the strand indentation. The knowledge gained can be used to design the strands for a certain bond behaviour.