Martin Prosser, Taylor Woodrow, UK
The proposed method of operation, where ballast moves under the sleepers by means of size and vibration from traffic, is indeed possible, and is the main problem with the use of stoneblowers and measured shovel packing-out of traffic conditions. During the overlifting a piece of regular-sized ballast will fall into the void, filling it completely and raising the level of the track above the intended design level once released from the lifting process. The offending singular piece of ballast is then subject to high loadings during the passage of the first few trains, causing it to disintegrate or forcing it into the stool of hardened ballast that supports the passage of trains. As the dislodged piece of ballast disintegrates or moves, this allows the line of the track to revert to its design profile supported by the full area of the stool.
The authors have cited a number of reasons why the two-layered ballast system should be adopted, but no discussion was entered into regarding the whole-life cost of the proposed system or the effect on maintenance requirements. If a value engineering study were to be undertaken on the proposals, there is no reason why the proposed benefits outlined could not be achieved by the inclusion of a vertical hole into the sleepers to act as a reviser of smaller stones, which would fill the voids created during the passage of trains and ballast/subgrade movement. The cost of installation of the proposed system would be greater than that of current track systems, owing to the increased complexity and the quality controls required to install the system.
The system as proposed by the authors would suffer from the use of packing and tamping machines, as these would mix the ballast grades as the tines are inserted and removed. Thus only measured shovel packing and stoneblowing would be desirable methods for correcting defects: this would create greater pressure on the limited machine resources and the discipline of contractors. The greatest problem the proposed system would cause is the effect on maintenance practices/resources. Any person who has undertaken measured shovel packing will know the proposed system will work and does behind every stoneblower except where the ballast is contaminated and therefore not free to move without being prised out of the finer material that has contaminated the ballast. This could be from the development of a wet bed or coal dust/sand dropped from above.
A defect in track level is not an urgent problem provided the defect affects both rails evenly. If a defect develops on only one rail this will cause the track to twist, which is a phenomenon that is known to derail trains of a specific length according to the severity of the twist. The development of twists would be the main problem with any track system that is designed to self-correct. The formation of twists would be particularly common where contamination of the ballast has occurred affecting only one rail, preventing the self-correcting action from working while on the other end of the sleeper the self-correcting action continues. This would require a proportion of the track maintainers' resources to be kept aside to complete urgent unproductive maintenance, thus increasing the overall life costing of the authors' proposed system. This may require a larger proportion of resources to be used for undertaking urgent maintenance than is required today.
Authors' reply
Our responses to the four points raised are as follows.
The risk of track uplift when small ballast is used. We investigated this possibility experimentally and reported on it in our previous paper.24 The system that we propose could be ‘fine tuned’ by the selection of ballast grading to achieve optimum levelling of the track without uplift. The historical review in the present paper indicated that small ballast has been used regularly in the past, and is currently used in other countries without uplift problems.
Costs. Our research project was part-funded by a manufacturer of concrete sleepers, and our discussions with them led us to the conclusion that any redesign of the sleeper would involve a far higher cost than a change in the ballast. We considered the provision of a vertical hole in the sleeper, but this would require repositioning of the prestressing wires, at substantial cost. Our studies indicated that our proposals would be cost-effective, but we were not able to arrange a site trial to prove this. We would welcome any opportunity to carry out a site trial.
Tamping and packing. Our paper concludes that the move towards larger ballast sizes has been driven forward by the requirements of tamping machines. We also conclude that this move has been detrimental to ride quality and maintenance requirements. We accept that tamping would be less successful with smaller ballast. We are suggesting that a strategic move towards smaller ballast maintained with stoneblowers would be beneficial.
Uneven support of the sleeper, causing twist of the rail. We see no evidence to suggest that our system would cause this. The arguments put forward to suggest that it might do so could equally be used to suggest the opposite. If a void formed under one end of a sleeper it would fill with small ballast particles and reduce the twist of the rail.
