Tuesday 18 February 2020
During August 2019, Iarnród Éireann undertook trial stabilisation works to the track at Corracullin bog on the border between Co Offaly and Co Westmeath, with the objective being to trial a method of track stabilisation without removing the tracks, writes Colin Hedderly.
The benefit of a more stable track is a lower deterioration rate of track geometry meaning less corrective maintenance is required by mechanical tamping machines.
Of greater significance in the long term, however, is the potential to bring about an increase in permissible speed in areas where the weak underlying subgrade causes the track to deflect under load.
The deflection of the track under a passing train results in a Rayleigh Wave propagating in front of the train. A speed restriction is necessary to ensure the train speed is kept below the velocity of the propagating Rayleigh Wave which is known as the ‘critical velocity’.
If the train were to exceed this ‘critical velocity’ it would generate large amplitude track deflections and have implications for safety.
Background
Studies show that bog covers around 21% of the surface area of Ireland which is more than any other country in Europe except Finland (Tanneberger et al., 2017).
The rail routes constructed through Irelands central plain involved construction of many miles across bogs (O’Dwyer and Cox, 2015). With depths of peat varying in thickness up to 20m (70 feet), to excavate or embank on these unstable masses was obviously impossible (Jennings, 1994).
The method of railway construction across bogs is attributed to G.W. Hemans and is described in his Account of the construction of the Midland Great Western Railway of Ireland, over a tract of bogs, in the counties of Meath and Westmeath published in 1851. Thanks to this record we have a good understanding of how railways were constructed through bogs in the 1840s and 1850s.
Construction involved dewatering the top layers of the bogs by excavating a system of horizontal and lateral drains along the intended line of the railway.
Once the drainage was working well and the bog formation subsided, two courses of heather sods were then laid to a width of 30ft on the consolidated bog surface, which had previously been given a profile rising to 18 inches at the centre, much in the manner of crossfall on a road (Cox and Gould, 1998).
The rails were bolted directly on to longitudinal timbers, which in turn formed part of a 25ft wide lattice framework bearing on the prepared bog surface.
The result was that no part of the track could deflect suddenly without fracturing, and both sides supported and counterbalanced each other (Cox and Gould, 1998). In some locations the track needed to be lifted by nine to twelve inches every day to make good the settlement that had occurred overnight.
This continued for many weeks and eventually settled down but vibration and differential settlement has always affected the track in these bog areas (Jennings, 1994).
In more recent times, the practice has been to increase the weight of the formation without causing formation failure. Experience was the guide rather than theoretical analysis.
Initially a 600mm deep mixture of limestone, chips and dust was laid in 200mm layers and above this was laid the ballast and track (Jennings, 1994). The conditions have stabilised but continue to require a high level of maintenance tamping.
The author is aware of a number of instances where formation failure did occur from overloading. The most notable of these being in April 2001 near Ballymote, Co Sligo, where failure occurred during tamping operations of newly relayed and recently ballasted track.
The track here was situated on a 3m-high embankment constructed over a peat bog. The failed section was reconstructed using lightweight polystyrene blocks to reduce load on the peat foundation.
Track bed stabilisation without removing the track
The engineering principle of piles acting as a steady support for structures built on top is long established and Iarnród Éireann has used piles previously to stabilise railway embankments.
In 2009 at Tubber bog in Co Galway piles were used in combination with a designed geogrid reinforcement to stabilise and support the track. To implement the design at Tubber bog it was necessary to remove the track and the work took several weeks to complete.
However, the uniqueness of the trial at Corracullin bog in August 2019 is that the piles were installed without removing the track.
The methodology to do this has been developed to allow the installation be carried out at night during 'engineering hours' and with minimal disruption to the normal operation of trains during the daytime.
Figures 4 and 5 illustrate the stress distribution in a simulated loading model before and after piling.
Trial pile installation works
Piling specialist Van Elle carried out the works using their remote controlled VE-SPIRIT mast drilling system. This is a fully caged rotary system with automatic cut-offs to maximise safety.
The piles installed were Van Elle’s Smartpile system which is based on screwpile technology. Each pile uses helical flight sections top and bottom to transfer load to deeper more competent layers.
Piles were installed in pairs in every sleeper bed at the ends of the sleeper. They were installed based upon a designed length to reach the more competent layers and a minimum required torque measurement correlating to pile load capacity.
The pile cap is kept 800mm below the bottom of the sleeper to be clear of any future track maintenance and renewal works.
Ground investigation was carried out in advance of the works involving window sampling and dynamic probes at regular intervals throughout the site.
This confirmed the existence of a substantial deposit of peat underlying the entire trial site of between 2.1 and 4.2m in thickness.
A conservative thickness of 4.5m for this layer was adopted by the designers in the pile layout design.
The ground investigation also confirmed that the ballast and track construction depths throughout the site were considerable at 1.2 to 1.5m in thickness.
This build-up in ballast depth under the track has accumulated over many years from the repeated correction of differential settlement by maintenance tamping and topping up with ballast.
In the short-term each top-up and tamp increases the trackbed stiffness modulus, but in the long-term the additional loads from topping up the ballast only contribute to the inducement of further track settlement which compounds the issue.
Analysis of track maintenance records confirm the site has been tamped seven times in the past six years which is a high level of maintenance intervention.
Atkins were engaged as design engineers to prepare design calculations and analysis of the buckling resistance of the piles.
The approach taken was conservative omitting all skin friction around the pile shaft because of the weak nature of the peat material and relying predominantly on end bearing resistance.
From the installation records, based on a correlation determined by Van Elle using CAPWAP (Case Pile Wave Analysis Program) and installation torque values, the piles achieved a vertical capacity ranging from a minimum of 147.7 kN to maximum 295.4 kN, with an average of 224.6 kN.
Measurement of improvement
In order to provide an immediate measure of improvement from the works, void monitoring equipment was installed on the track for a month prior to, and a month after installation of the piles.
The equipment was called VoidMate and supplied by Garco Ltd. It provides a continuous recording of local deflections at track level under train loadings.
The VoidMate unit is clamped to the rail and operates by measuring the vertical displacement of a plunger placed directly upon or within the upper ballast layers. See image of VoidMate equipment in figure 8.
There were a total of six void monitors installed within the trial site. The results from the two void monitors located at the centre of the trial site area show a clear reduction in vertical track deflections following installation of the piles.
The average reduction in deflection in both cases is approximately 50%. Figure 9 is the summary results for one of these monitors (ref. 22029) which shows a reduction in average deflection from 5mm to 2mm. The reduction in maximum deflection is from 25mm to 2mm.
Summary
The trial achieved its objective by demonstrating a methodology of stabilising the ground through piling without having to remove the track and disrupt train services.
The results of the trial are very promising and clearly show an immediate significant reduction in deflection magnitude, whilst the installation method created little disturbance to the track.
Further assessment of the trial site using Falling Weight Deflectometer equipment is hoped for in the future. This is another method to assess the support stiffness to the sleepers.
A comparison of data from the test site with the adjacent non-piled track would provide another measure of the improvement.
In all 146 No. piles were installed in the trial with valuable experience gained in the logistics of undertaking this type of work.
There are areas where further efficiencies will be gained, but the installation time of 9 minutes for a pile to a depth of 7m was typically achieved during the works and was deemed efficient.
One potential concern following the trial was whether the change in trackbed stiffness would be sudden and noticeable from a passenger comfort point of view. This has not shown itself to be an issue of concern.
The way the vertical load distribution spreads out below the sleeper combined with the pile cap being kept 800mm below the bottom of the sleeper provides for a transitioned change, and this was proven by the VoidMate monitors situated at the beginning and end of the trial site recording a reduced improvement in deflection magnitude.
As to the future and the authorisation for more of this work it will come down to a favourable cost to benefit analysis and funding.
Individual sites will have to be assessed for their benefit and in particular the opportunity of higher speeds and journey time improvement that can be delivered by it.
There is further work to be done here but in principle it has been demonstrated that there now exists an engineered solution that can be targeted to stabilise areas of weak underlying subgrade without interfering with the running of trains.
A short video of the installation work can be found here:
[embed]https://youtu.be/nrB5EFd-ZAM[/embed]
Acknowledgements
The author acknowledges the support from all colleagues who worked on the project in Iarnród Éireann, and also to Van Elle and their John Allsop and Martin Gregory especially.
Author: Colin Hedderly is a senior track and structures engineer with Iarnród Éireann. He is a chartered engineer with more than 20 years’ experience in the rail industry in Ireland and the UK and is the lead engineer for the western division of the chief civil engineer's department, which has responsibility for maintenance and renewal of the track and structures.
References
1.) Aalen FHA, Whelan K, Stout M (1997) The Atlas of the Irish Rural Landscape. Published by Cork University Press.
2.) Cox RC, Gould MH (1998) Civil Engineering Heritage Ireland. Published by Institution of Civil Engineers.
3.) Hemans G (1851) Account of the construction of the Midland Great Western Railway of Ireland, over a tract of Bogs, in the Counties of Meath and Westmeath. Transactions of the Institution of Civil Engineers of Ireland, vol. 4.part 1. 1851. pp 48-60.
4.) Jennings PO (1994) I’ve been Workin’ on the Railway. The Institution of Engineers of Ireland Presidential Address.
5.) Musgrave P, Wehbi M, Jackson L, Stevenson A, O’Neil L, (2017) A Guide to Trackbed Micropiling. Published by Network Rail IP Track Bed Design & Innovation Group. See https://www.thepwi.org/technical_hub/technical_hub_files/a_guide_to_track_bed_micro_piling
6.) O’Dwyer D, Cox R, (2015) Early Irish Railway Construction. The proceedings of the Second Conference of the Construction History Society
7.) Powrie W, Le Pen L (2016) A Guide to Track Stiffness. Published by the Cross Industry Track Stiffness Working Group
8.) Tanneberger et al. (2017) The peatland map of Europe. Paper published by the International Mire Conservation Group and International Peatland Society.