2020 Heavy Haul Session
Research into the Causes of Visually "Undetectable" Broken Spikes
University of Illinois
(Retired) Norfolk Southern
Multiple wide-gauge derailments have recently been attributed to broken spikes that were not identified prior to the event. These failures have been primarily found in premium elastic fasteners, a type of fastening system that uses an elastic clip to hold the rail to a tie plate and spikes to hold the plate to the crosstie. Premium fasteners have gained popularity in North America in heavy axle load (HAL) freight service for reducing rail-rollover risk, lowering maintenance costs and generally not needing rail anchors. However, broken spikes in these systems have become a safety concern, especially considering that most failed spikes are visually undetectable. Additionally, the life-cycle cost of elastic fasteners is increased due to labor-intensive inspections and premature component replacement. For this reason, a project at the University of Illinois at Urbana-Champaign funded by the Federal Railroad Administration is investigating loading environments that lead to spike failure, and the response of premium fastening systems.
This presentation provides an overview of field data acquisition of two tracks in the same curve in revenue service on a North American heavy axle load (HAL) railroad. One of the tracks sees mostly traffic of empty trains going uphill and has high occurrence of broken spikes, while the other track has primarily loaded trains with airbrakes applied going downhill and has fewer spike failure occurrences. Since both tracks are located in the same curve with constant curvature, degree, and climate, the experiment allowed the researchers to isolate the effects of weight, speed, direction, temperature, etc. in the vertical, lateral and longitudinal loads applied to the rails and, therefore, to the fastening systems. Preliminary analysis of the data captured is presented. Findings from this work will be used in laboratory experiments, compared with their results, and in the validation of finite element models also in development, while simultaneously being used to further the industry’s understanding into longitudinal load transfer.