Areview of the geotechnics of railway track was presented by Dr Phil Sharpe of Scott Wilson Pavement Engineering (SWPE).Starting with the basic structure and function of each layer of the track bed, he highlighted problems that can occur within the track support structure and described techniques to investigate and predict such problems.
Track support structure The track support structure may be divided into a number of layers, each with a specific role.The sub-formation layers are designed to provide support to the ballast which supports the track.Figure 1 shows a complete track bed for a modern highspeed line.This is the optimum structure, although on older lines ballast support is provided by some or none of the subformation layers depending on location, age and loading of the track bed.The aim of sub-formation design is to ensure that ballast is the only layer that should require maintenance to maintain good track quality.
Railtrack assesses track quality by measuring the standard deviation (SD) of the level profile about the mean, over sections one-eighth of a mile long (allowing for vertical alignment).On a new high-speed line an average SD well below 1mm can normally be achieved, but on UK tracks it is often difficult to achieve an SD of 2mm. Railtrack requires that 90% of a 125mph route is maintained to an SD better than 2.4mm.
Other than its primary function to support the track, ballast is required to be adjustable (to maintain track level) free draining possess good load spreading properties to dissipate applied track stresses to the sub-formation and act as a fines reservoir to collect fine particles which form during the life of the track bed.
The only routine track maintenance required should be regular tamping to re-compact and reposition ballast, thereby maintaining track quality and alignment.Tamping is typically performed using specialised track-mounted plant and is usually performed annually, or more frequently if the track bed is approaching the end of its life.
Problems with track bed Ballast deterioration During the loading experienced in the track bed during the passage of a train (cyclic loading incorporating rotation of principal stresses) and during tamping, the ballast particles grind together to produce fines which eventually fill and choke the voids between the single size ballast (50mm).With infiltration of water the fines in choked ballast can form slurry beneath the rail sleepers and this leads to a reduction in track support.This is frequently evidenced by splashes of slurry on the sleepers and rails.To remedy this, ballast must be cleaned and rail-mounted plant is used to remove and sieve the fines from the old ballast and replace the cleaned material.This is typically required every 15 years.
Pumping Water trapped at the interface between open textured granular material and a cohesive layer causes cohesive materials to liquefy, forming slurry which progressively pumps up into the ballast, again affecting track support and resulting in erosion of the cohesive layer.This is remedied by providing a sand blanket beneath the ballast of between 100mm (with a geotextile) to 300mm thick to aid drainage and restrict the passage of the fines.
Cess heave This is a bearing capacity type failure of the track bed.Following wetting or dissipation of negative pore water pressures in the subgrade, the weak subgrade is forced up into the cess (Figure 2). Additional ballast compacted beneath the sleepers to maintain track support tends to increase the rate of cess heave.This type of failure used to be typical on clay soils, in particular on overconsolidated clays, but it is now uncommon.This type of failure is typically repaired by increasing overall track bed depth, for example by using a sand blanket, to improve drainage and increase frictional resistance on the shear surface.
Critical velocity The critical velocity of the soil is a function of the speed of wave propagation through the material.For a stiff soil this is typically 300-400mph, and for softer soils 60-70mph.As the train speed approaches the critical velocity, resonance occurs and track deflections increase dramatically (in some cases by more than 100%).
Earthwork failure The investigation of earthwork failures and design of remedial measures is considered to be a separate discipline from track bed failures, although they clearly interact. Slope failures encompassing and undermining the track bed are normally investigated and repaired using standard geotechnical techniques.Earthwork problems must be addressed in order to ensure high performance from the track bed, although it is equally important to note that stable earthworks do not necessarily guarantee good track quality.
Track bed investigations Traditional track bed investigation techniques involved the excavation of shallow trial pits (400mm deep and 100m apart) and visual inspection to assess track bed condition.For deeper investigations, hand augered trial holes were used. However both these techniques present problems on live tracks.More recently, following an extensive period of development and testing, three new techniques have been introduced by Scott Wilson Pavement Engineering.
Ground penetrating radar (GPR) GPR is used to assess the thickness of the ballast layers with measurements being made in each sleeper bay. This gives a long section/profile of the base of the ballast and this can be compared with the track quality profile. The GPR technique is extremely quick and can be used as a preliminary assessment of deterioration modes (ie where areas of poor track quality and track bed anomalies coincide) allowing identification of thin ballast/subgrade pumping or wet spots.This technique can therefore be used to find the extent of problem areas and target further invasive investigations.
Automatic ballast sampler (ABS) The sampler contains a plastic tube which is hand driven into the track bed to obtain a sample/vertical profile for inspection in the laboratory.Sampling at regular intervals enables a longitudinal profile of the track bed to be made.A classification system has been developed to allow assessment of the samples collected.The system is based on a visual inspection of grading to assess ballast deterioration and permeability to assess cleanliness/subgrade erosion (Figure 3).The ABS is a rapid technique which enables a detailed assessment of track bed and subgrade conditions in problem areas. It allows evaluation of ballast depth to enable tamping and ballast cleaning to be performed to appropriate depth to prevent disturbance of the subgrade/sub-formation.This data can also be used for more complex analysis discussed below.
Falling weight deflectometer (FWD) The FWD is a device commonly used for the evaluation of road pavements.A modified version is used to assess rail track.An impulse load is generated by dropping a large weight via a rubber buffer on to a beam to load a sleeper that has been released from the rail.Geophones are mounted on the sleeper and ballast to measure deflection (Figure 4).
The geophones measure different components of the deformation induced by the loading, as shown on Figure 4.The geophones on the sleeper measure the deflection of the top of the ballast under load (D 0), the closest ballast geophones are considered to measure the deflection of the base of the ballast (D 300 ), and the furthest the sub-formation component (D 1100 ).Typically a third of the total deflection is caused by each element.From deflection and load data the stiffness of the track bed can be found. Comparing FWD and track quality data has shown track quality is a function of stiffness as well as other factors such as ballast and sub-formation condition.A relationship has been developed where track quality (TQ) has been shown to be a function of FWD stiffness (E) normalised for the grading coefficient of uniformity of the ballast (U) assessed from materials collected from the ABS.This relationship, which includes a measure of ballast condition, allows track quality to be predicted.
TQ = f(U, E) In addition, load time data collected by the FWD can be analysed to assess the speed of the induced deflection pulse through the track bed and hence the critical velocity can be established. Research has shown that where line speeds are 30% less than critical velocity acceptable performance is observed.Therefore the FWD can be used to assess the suitability of (or even set) speed limits.
A trial on a very soft subgrade site at Purfleet in Essex has shown that by assessing critical velocities with the FWD, and installing a reinforced track bed to increase stiffness and hence critical velocity, the speed limit could be raised from 60mph to 90mph.Line speeds could be also be improved in areas of low critical velocity by using ground improvement techniques to increase track bed stiffness.
Conclusions Track bed deterioration mechanisms are well understood, and the influence of track bed stiffness is now appreciated.Stiffness together with material condition assessment has been incorporated into a model of track bed deterioration which can be used to predict track bed quality.A suite of new track bed assessment techniques has been developed and a combination of these two elements can be used to manage and optimise track bed operation, renewal and repair.
If further routine track bed investigations are performed with a programme of complementary targeted track bed improvements and remedial works, this should contribute to reduced expenditure on track bed maintenance in the longer term.
Discussion Sharpe added that the investigation techniques reported had been accepted by many Railtrack zones, and he hoped they would be incorporated in Railtrack guidance notes for track bed investigation.
Commenting on track bed improvements, he said on new high-speed lines the high stiffness was provided by having large construction depths, for example, 1m of prepared subgrade was commonly laid, before placing any track bed layers.The highway specification for earthworks provided an adequate construction platform upon which to construct rail track and adequate drainage was equally important.
Unfortunately, UK rail earthworks were not generally built to such high standards and often had poor drainage.
Thorough investigation using the techniques described, followed by correct design and construction of remedial measures, would enable permanent improvements to be made to track quality.
Note: All figures are copyright Scott Wilson Pavement Engineering.