Behaviour of composite sandwich panels bonded with epoxy polymer matrix for railway sleepers

Behaviour of composite sandwich panels bonded with epoxy polymer matrix for railway sleepers

The Australian railway industry spends millions of dollars every year in replacing poor condition railway sleepers in order to maintain the track quality and ensure a safe track
operation. It is estimated that more than 2.5 million timber sleepers per year are required for railway track maintenance. Over the last decade, it has been increasingly difficult to get suitable quality hardwood to keep up with the demand for railway maintenance. Moreover, the global
environmental impact due to felling huge number of trees for manufacturing timber sleepers is a major concern. In recent years, significant efforts have been provided towards the development of polymer composite sleeper alternatives to replace deteriorating timber sleepers. Despite their potential, the uptake of polymer composite sleepers has been extremely limited because of the high production cost of these technologies. An improved understanding of composite sleeper technology such that improved designs and effective material usage needs
to be developed to reduce the overall cost of production.

This study focused on investigating the behaviour of a fibre composite railway sleeper manufactured from composite sandwich panels and the panels are bonded together with epoxy polymer matrix. An intensive characterisation of the epoxy polymer matrix and composite sandwich panels was conducted to determine the effective usage of these materials for railway sleepers. The effect of resin-to-filler ratio on the thermal, physical, mechanical and durability properties of the epoxy polymer matrix were evaluated. Filler materials composed of fire retardant filler, hollow microsphere and fly ash were increased from 0% to 60% in the epoxy based matrix. The results showed that epoxy-based polymer mixes containing 30% to 50%
fillers by volume provided a good balance of thermal, physical, mechanical, and durability properties suitable for coating the composite railway sleepers. The capacity of single composite sandwich beam in resisting bending and shear forces were then determined by testing them in
horizontal and vertical orientations and with a shear span-to-depth ratio varying between 0.5 and 12. The results showed that the horizontal sandwich beams performed better under bending while vertical sandwich beams were more effective in resisting shear. It was also found that
the beam orientation has more influence on the load carrying capacity and stiffness properties than changing the shear span-to-depth ratio.

The structural integrity and composite action between the sandwich panels and epoxy polymer matrix were investigated. The effects of binder properties, bond length, bond thickness and bond width were investigated to evaluate the bond behaviour of a composite sandwich panel and epoxy polymer matrix. The results indicated that the bond thickness has the greatest influence on the bond strength followed by the properties of polymer matrix, bond length and bond width. A bond thickness of 5 mm, bond length of 70 mm and using polymer matrix with 40% filler and 60% resin (by volume) can eliminate failure in the glue line and promote failure in the sandwich panels. Using these bond parameters, full-scale layered sandwich beams were
manufactured and their structural behaviour was evaluated under four-point bending and asymmetrical beam shear tests. Results showed that the binding of sandwich panels using
epoxy polymer matrix can prevent the wrinkling and buckling of the fibre composite skins as well as the indentation in the phenolic core. This concept increased the bending and the shear strength of the vertically oriented beams by as much as 25% and 100%, respectively, compared to single sandwich beams. Using the same amount of material, the vertically layered beams exhibited similar bending strength and 50% higher shear strength but only 7% lower effective
modulus of elasticity compared to horizontally layered beams. More importantly, the layered sandwich beams were found to have strength and stiffness similar to the hardwood timber.

The optimal design of layered sandwich beams for railway sleepers and the performance evaluation under static loads were conducted as the last study. The optimal shape
of sleeper under quasi-static load was obtained using topology optimisation. The optimal sleeper shape requires only 50% volume of materials compared to a rectangular timber sleeper. Moreover, the rail seat and centre bending moments, shear strength, screw holding capacity,
and electrical resistance of optimised composite sleeper are higher than the traditional hardwood timber and exceed considerably the performance requirements for a railway sleeper. The handling, installing and fastening system of this composite sleeper were similar to timber and require the same equipment and machineries. A total of 50 sleepers have been installed on a trial basis in the Southern line rail track at Nobby, Queensland, which are performing very
well and are expected to outperform its design life.

An in-depth understanding of the behaviour of a new type of composite sleepers made from layered sandwich panels bonded by epoxy polymer matrix was the significant outcome
of this study. The experimental data, simplified theoretical models and finite element simulations derived from this study can be used as important tools for a safe design of composite railway sleepers. The optimal sleeper shape proposed in this study can provide a cost-effective alternative to timber sleepers in a mainline track to decrease railway track maintenance, and provide a totally recyclable and sustainable sleeper technology.