BALLASTLESS track is an attractive option in situations when ballast is scarce and a low-maintenance solution is required. In desert conditions, ballast fouling by wind-swept sands is often so severe that the ballast is unable to perform its essential functions and the performance of the track structure is seriously compromised.

The use of ballastless track structures in desert conditions originated from a critical evaluation of the performance of traditional ballast track structures in areas where severe fouling by wind-borne sand occurs frequently or continuously.

Conventional track incorporates material to create a maintainable and durable structural layer between the sleeper and the sub-ballast. In desert conditions, wind-blown sands contaminate the track structure by entering the ballast as well as the wheel/rail interface. The effects of particulate contamination in machine-element contact areas, of which the wheel/rail interface is one example, have been studied in detail. Contaminated interfaces by particles varying in size from nanometres to micrometers can be responsible for increased wear and for even more severe failures.

Sand fouling ballast has a detrimental impact on its function. In most cases, the construction of sand barriers or the lifting of the entire track structure has been unsuccessful in keeping the sand away and the only solution is to employ an alternative track structure, one possible option being ballastless track.

Various ballastless systems are in use around the world, and most are slab track systems. These can be divided into two sub-categories: those with discreet rail support, with rails hand-laid into reinforced concrete slabs fitted with baseplates (Rheda 2000, Sonneville Low Vibration Track, Japanese reinforced concrete roadbed system), and systems with continuous rail support such as Paved Concrete Track (Pact) and Embedded Rail Structure (ERS).

Another example of a ballastless track structure that incorporates continuously-supported rails is Tubular Modular Track (TMT), which has been under development in South Africa since the late 1980s. Modules approximately 6m long are pre-cast, and the gauge is set by galvanised steel bars at a spacing that is determined by the application and design requirements. The rails are continuously supported on longitudinal reinforced-concrete beams.

Tubular track is currently being tested by the University of Pretoria and has also been evaluated by the Transnet Freight Rail (TFR) Track Technology Centre.

Finite element analyses (linear transient dynamic models) were used to determine the formation life that can be expected with the use of TMT. Factors influencing this calculation included estimated traffic levels and axleload. An 18m-long 3D brick model was created representing three TMT modules, with a width of 4m and a depth of 3m. The formation consists of four 200mm layers. The load on the structure was the static wheel load of 15 tonnes, multiplied by a load factor of 1.34.

The formation design life prediction was based on the South African mechanistic pavement design analysis method. In the top granular layers, the life prediction depends on the stresses in the three major directions in the middle of the layer. In the non-granular or subgrade layers, the life is based on the material's vertical strain at the top of the layer. The load repetitions applied to each layer are calculated and the layer with the lowest load repetitions is the critical layer.

Tubular-tableA comparison between conventional track and TMT systems indicated that stresses are higher in TMT, but deviation stresses are lower, resulting in a longer life prediction. Analyses also showed the lower layers are the critical ones.

In the case of the TMT installation on the Riyadh - Hofuf line in Saudi Arabia, the life prediction was 38 years, based on an initial traffic level of 25 million gross tonnes per year and a 10% annual increase in volume.

Tubular-fig1 2
Laboratory tests were carried out on a full-scale TMT test section at TFR's Track Technology Centre. The modules, designed for an axleload of 22 tonnes, were placed on two different formations, namely a bitumen-stabilised and a cement-stabilised foundation.

With dynamic, repetitive loading at three different testing positions, it was found that TMT performed well when subjected to five million load repetitions at a 22-tonne axleload with a design life of 20 years. A dynamic factor of 1.6 was used during the loading. Figure 1 shows the permanent deformation of the track section at mid-beam (Position A), at the transition (Position B) and at mid-beam for fully-saturated conditions. Even after increasing the magnitude of loading, failure did not occur.

The results indicated that TMT would be suitable for well-designed subgrades, provided that resilient pads are used to reduce the high dynamic impact forces.

Field tests

Field tests were conducted on a 120m section of TMT at Amandelbult on the Rustenburg - Thabazimbe line in South Africa, which has a maximum axleload of 20 tonnes. The section was instrumented with strain gauges, multi-depth deflectometers, and remote video monitoring (RVM). A section of conventional track with similar formation conditions was also instrumented at the same time. Tests were conducted on this section in 2004-05, and 2010. Table 1 shows a comparison of the deflection measurements during both test periods, and typical results obtained in 2010 with RVM instrumentation are shown in Figure 2.

The Amandelbult tests revealed that ballast deflection has a high influence on the total track structure and does vary over time, as Table 1 demonstrates, whereas the TMT structure deflection remains constant at the mid-beam and increases slightly at the joint between modules. Beam displacement measurements at four locations showed that the ends of the adjacent beams do not move vertically in opposite directions during dynamic loading. Field tests with an instrumented wheelset showed that wheel/vehicle accelerations were lower on tubular track than conventional track.

During 2009, field testing with a test train was carried out at Centurion station near Pretoria on TMT and conventional track using RVM to measure rail deflections. Comparative measurements were taken to evaluate the effect of speed on resilient deflections in both of these track structures. On the conventional track, deflection of the sleeper and formation were measured, while beam and formation deflection on the TMT section were measured in the middle of the 6m-long concrete module.

tubular-fig-3The deflection results presented in Figure 3 relate to an axleload of 15.1 tonnes at speeds ranging from 10km/h to 40km/h. The absolute formation deflection on conventional and Tubular track was approximately the same - roughly 0.35mm - so both types had a similar effect on the track formation. But the absolute total track deflection of conventional track was twice as large as the absolute total track deflection of the TMT modules. On conventional track at a speed of 30km/h, the sleeper deflection was 1.15mm while the beam deflection was 0.57mm on Tubular track under the same conditions.

TMT has been installed at a number of other locations where it operates under heavy haul and/or desert conditions. During 2006, four 1:12 points were installed at Ermelo yard in South Africa. These carry slow-moving traffic at an axleload of 26 tonnes and have carried around 500 million gross tonnes since they were laid. Compared with conventional turnouts in the yard, rail wear and general maintenance have been noticeably lower on the TMT sets. The stiffness of the TMT modules limits movement of the track components and reduces overall maintenance.

A 25km TMT section was installed on the 1065mm Aus - Lüderitz line in Namibia in 2005. The absence of local ballast sources in this desert area, combined with the presence of shifting sand dunes and the narrower formation width have made this an ideal site for the installation of ballastless track.

In Saudi Arabia, a 1km test section was constructed on the Riyadh - Hofuf line, which carries 5 million tonnes of freight per year with a maximum axleload of 30 tonnes. Once again TMT has been an economical solution in an area where ballast sources are scarce and where wind-swept sands contaminate the ballast, destroying its essential properties.

Extensive research and development over the last decade has contributed to the success and performance of ballastless track structures. A wide range of experimental tools has been used to evaluate TMT in South Africa, both in terms of its structural design and performance in the field. The results of these tests confirm the ability of TMT to deliver the same, or in most cases, better performance than conventional track.

In conclusion, the research identified the most significant advantages of TMT over conventional track as being:

• lower formation stresses due to the larger effective contact area

• reduced track and rail deflection resulting in longer component life

• fewer maintenance requirements in terms of geometry correction

• reduced earthworks volumes due to the narrower track structure, and

• improved vertical and lateral stability.

It is believed that the results of these investigations will enhance the competitiveness of rail transport in South Africa and elsewhere through the development of more effective and efficient track infrastructure.

This article is based on a technical paper by Hannes Gräbe and Jaco Vorster, University of Pretoria Chair in Railway Engineering; F J Shaw, Transnet Freight Rail (Track); and S C A Van Haute, VKE Consulting Engineers, South Africa, presented at last year's International Heavy Haul Association conference in Calgary.