Water - the most overlooked component of low adhesion

Each year low adhesion causes significant disruption across the UK rail network and ultimately costs rail users up to £350m. The best-known manifestation of low adhesion occurs each Autumn when fallen leaves are compacted onto the rail head by passing trains, creating a hard, “Teflon-like” coating on the rail head. When the leaf-based layer is lightly wetted, the resulting poor wheel-rail friction leads to greatly increased braking distances and reduces starting traction. This phenomenon, often ridiculed in the media as “leaves on the line”, causes major disruption across large parts of the network. Low adhesion also occurs at other times of the year in the presence of rail head contamination (such as iron oxides) and light rain or dew. The common thread running through all these examples is the presence of a contaminant and a certain critical amount of water. While a dry rail and a fully wetted rail give acceptable levels of friction, a damp or lightly wetted rail is associated with poor adhesion. Many train drivers felt water might play a part in low adhesion but there is now a substantial body of research which confirms and quantifies the issue. Here is a summary of this work.

Scientists working at British Rail Research as long ago as the 1970s conducted rig-based studies into the effect of water on the friction between two steel surfaces. In one experiment [1], a slow constant flow of water was applied to a rail/wheel simulation rig – graph shown in figure 1.

As expected, there was an immediate reduction of the friction coefficient which was maintained by a slow flow of water (μ>0.5 dry to μ=0.2 damp rail). A large bulk of water was then applied to the contact point which gave an instantaneous increase in the friction coefficient (μ=0.2 damp to μ=0.3 wet rail). Next the water supply was turned off completely, allowing the contact point to start to dry – this resulted in the friction coefficient falling again to its lowest level (μ=0.3 wet to μ=0.1 nearly dry rail). Finally, as the surface became completely dry, the friction coefficient returned to its initial high level (from a low of μ=0.1 for a nearly dry rail to μ > 0.4 dry rail).

In another study [2], water was mixed in a test rig with wear debris and surface rust to form a lubricating paste with resulting friction coefficients as low as 0.05. The lowest friction values occurred when the ratio of water to debris was small and a viscous paste was formed; this was observed for a few seconds whenever the wetted friction surfaces were on the point of drying completely. The paste was described by the researchers as non-Newtonian – i.e. a fluid whose viscosity varies with shear rate.

Figure 1: Effect of water on friction

Figure 2: Comparison of adhesion levels on sprayed and un-sprayed tracks

The findings from these studies and observations of the positive effect of heavy rain on rail head contamination encouraged researchers to test the effect of water in the real world. In 1979, trackside water spraying equipment was tested at Bearsted Bank (near Maidstone, Kent). The adhesion conditions in Autumn were known to be especially bad at this location with several leaf fall-related braking and traction issues recorded. Rail friction values were collected for a sprayed section of the track and a “control” section which remained untreated [3]. Figure 2 shows the results of these tests.The results showed a statistically significant improvement in friction with water spraying. The lowest mean adhesion recorded on the treated section of track was 0.11 while the lowest value for the control was 0.04 – this value being low enough to cause major problems with traction and braking.

While the evidence from these tests confirmed the benefits of trackside spraying, it has never been implemented on the UK rail network, perhaps due to concerns about the cost and complexity of such a solution.  

Over the following 30 years, with a few exceptions [4, 5], the spotlight moves away from water. Instead, research concentrates on sanding systems (one-shot and automatic), wheel slide protection, Laser rail head treatment (under the LaserThor trade name) and the impact of low adhesion on the proposed European Rail Traffic Management System (ERTMS). Arguably, only limited progress is made towards solving the issue of low adhesion during this time.

In 2016, The University of Sheffield undertook a study funded by RSSB: Modelling and quantifying the influence of water on wheel/rail adhesion levels. During this project, adhesion tests were run using lab-scale rigs and at full-scale using the tram wheel/rail test rig at The University of Pardubice in the Czech Republic. These tests once again showed that the adhesion level was only marginally reduced by large amounts of water in the contact, while a significant reduction of the adhesion level was observed when applying low amounts of water.

Figure 3: Creep curves for different water delivery rates and improvement as water rate increases

Figure 3 shows a sample of the creep curves generated using a water droplet application system at a rolling speed of 5 m/s. At very low droplet rates (25 μl/s) the level of adhesion is very low. As water droplet rate is increased to 35 μl/s there is a clear change in adhesion behaviour. At the low water rate there was a rapid decrease in adhesion upon sliding. The traction control system on the rig was unable to effectively return to pure rolling once sliding was initiated and high creepage rates were seen.

Levels of adhesion were reduced to below 0.1 (T/N). The highest water delivery rate (350 μl/s) produced similar curves to those bulk water application tests. These tests once again indicate that low adhesion only occurs over a small range of mixture proportions – i.e. when a critical amount of water is present. The tests also quantify the amount of water required to generate low adhesion and confirm the minimum water delivery rate needed return adhesion to a “safe” level.

As a result of the research undertaken over the last 50 years, the role water plays in low adhesion is now very clear and quantifiable. The Water-trak system aims to exploit this knowledge by delivering water to the wheel-rail interface when low adhesion conditions are detected, creating step-change improvement in braking and traction.

 References:

1. A basic study of wheel/rail adhesion laboratory studies of the effects of Water – T.M. Beagley & C Pritchard, Tribology Section, Railway Technical Centre, 1973

2. T.M. Beagley, C. Pritchard, Wheel/rail adhesion - the overriding influence of water, Wear, 35(2), 1975, pp.299–313

3. Taylor R.K, British Rail Research, The effectiveness of trackside water sprays on leaf affected track at Bearsted Bank 1979 (TM-TRIB-43)

4. Rail Contamination and It’s Influence on Wheel/Rail Adhesion, IJ McEwen, BR Scientifics, January 1999 (page 69)

5. T354 Review of low adhesion research, conducted for RSSB by C R Fulford Associates 2004

6. T1077, Modelling and quantifying the influence of water on wheel/rail adhesion levels – Phase 2 report, conducted for RSSB by University of Sheffield, L Buckley-Johnstone, University of Sheffield, R Lewis, University of Sheffield, K Six, Virtual Vehicle, G Trummer, Virtual Vehicle, 2016