Safety Issue
The track had been maintained based on the assumption that rail
creep is sufficiently homogenous in open track to be manageable if
monitored at 500 m intervals. It is likely that the rail creep in
the area of the derailment was not homogenous as a result of long
term and recent track work. Ballasted track structure based on
continuous welded rails is sufficiently non-homogenous to enable
bunching points to develop over time in open track. If localised
rail bunching occurs, a build up of compressive longitudinal forces
in the rail, sufficient to cause a misalignment, can occur over a
short distance. Undertaking track work that breaks the bond between
the sleepers and ballast can result in localised creep and bunching
of rail. The current approach to creep measurement, that is,
monitoring rail movement every 500 m, may not detect these
localised variations.
Safety Advosory Notice
The Australian Transport Safety Bureau advises that all track
maintainers in Australia should consider the implications of this
safety issue and take action where considered appropriate.