Geotechnical engineering is vital to a railway system. Most railways are built on the ground, and for much of their length run within or on geotechnical structures such as cuttings and embankments. The principles of soil mechanics, which underlie the practice of geotechnical engineering, may equally be applied to a ballasted trackbed.
A core aim of the research work is to optimise track design and reduce the need for tamping maintenance (shown in Figure 1). Tamping is a process by which the track is lifted to its design alignment and the ballast is squeezed beneath the sleepers using vibrating tines, to restore the geometry. This is necessary because loss of geometry is similar to the formation of potholes in roads: it leads to poor ride comfort and, if left unchecked, ultimately becomes a danger to vehicles and passengers. However, while tamping restores geometry it also damages the ballast by crushing particles and generating finer fouling material, which adversely affects drainage and performance and hence the residual life of the ballast. Some 10 or 15 tamping cycles may take place over 30 years before ballast replacement. The interval between tamping cycles reduces with each one so, depending on the line use, tamping might become uneconomic when the interval between required tamps to maintain a certain quality of geometry reduces to less than, say, a year. The use of tamping should therefore be limited as much as possible to extend ballast life.
The tamping machine extends many tens of metres and operates a number of tine banks. Tampers usually operate during night time when tracks are not in use and can cover some kilometres of track during a single shift.
Tamping frequency may be reduced by applying incremental improvements to the track system as shown in the chart (Figure 2) such as using under sleeper pads, strengthening ballast with random fibres, modifying the ballast gradation, using stoneblowing instead of tamping, and more. Many potential improvements are the subject of ongoing research at Southampton to quantify their relative benefits in relation to cost.
.jpg_SIA_JPG_fit_to_width_INLINE.jpg)
Certain parts of the rail network account for disproportionate costs. Examples include the transitions from normal ground onto hard substructures and complex track geometries such as switches (points) and crossings. Between them these locations account for less than 1% of route length in the UK, but about 20% of maintenance costs. Investigating the performance of track designs at such critical zones has the potential to lead to improved designs and reduced maintenance costs in the future. Geophones are being used to give insight into real track behaviour at such zones. These are small seismic sensors that can be mounted on sleepers and the signals processed to determine the frequency content and displacement of the track as a train passes. Figure 3 shows an instrumented section of track at the approach to a level crossing.
Other common problems on the track include wetbeds ( Figure 4 ) which lead to progressive failure of local sections of track. Such zones often occur on soft clay/silt formations and are evident by the appearance of pumped clay or silt on the ballast surface, as shown in Figure 4 . Monitoring of such sites is being carried out to assess the effectiveness of track renewal procedures in preventing their re-occurrence.
Laboratory testing of scaled ballast using Triaxial apparatus (Figure 5) is used to determine material properties which can then be related to performance in large scale tests and in service. It can also be used to validate numerical models and calibrate models that are used to predict the performance of larger systems.
Ballast behaviour can be simulated and improvements investigated theoretically using advanced numerical modelling techniques. We employ a unique potential particle based discrete element method (DEM). Modelling software was developed in-house to make very efficient use of the University's world standard supercomputer. The interactions between each individual particle of ballast and the track it supports are captured, analysed and assessed in order to predict the behaviour of the system and understand the influence of different factors on that behaviour. ( Figure 6 )
A representation of a single sleeper bay of track is used to evaluate the relative performance of different sleeper and ballast combinations including the use of under sleeper pads ( Figure 7 ).
On the West Coast Main Line (WCML), Pendolino trains with a tilting mechanism operate at higher cornering speeds than ever before. Maintenance staff have observed sporadic occurrences of local ballast migration due to the high cornering forces developed. Ballast migration is a process whereby the ballast gradually moves down the super elevated (canted) track and gathers in a heap against the low rail. Field measurements using geophones are being used to gain insights into the mechanisms by which the ballast migrates, with the aim of better understanding the mechanisms involved in order to design preventative measures. Figure 8 shows a Pendolino approaching a ballast migration spot during the snowfall events in early 2013. Figure 9 shows a close up of a ballast migration feature at a different site.
New research into principal stress rotation in unsaturated soils reflects real conditions better than the idealized loadings currently used for design. Our work will help engineers predict long term degradation of the sub base in relation to traffic loading, plan maintenance and develop cost-effective remedial measures. New data will also input to an improved track system model. See figure 10 .
Ballast fouling adversely affects drainage, increasing the rate of track geometry degradation and reducing the effectiveness of tamping. Our high powered computed tomography equipment is being used to develop a sound scientific understanding of the mechanisms involved, to enable the design of more resilient track. ( See Figure 11 )
