Cyclic Loading Tests of Concrete Sleepers With Varying Ballast Condition

Abstract

The experiences on the use of concrete sleepers in Finnish railways are mainly very good when the supporting layer, ballast, consists of coarse-grained crushed rock aggregate. Nowadays in Finland there are still wooden sleepers on many lines carrying little traffic. On these lines the ballast often consists of gravel providing essentially weaker support for the sleepers. One question in future renovations is whether concrete sleepers remain undamaged with gravel ballast. To test the bending behaviour of four different sleeper models resting on three different ballast beds, a cyclic loading test arrangement was built up in the laboratory of the Department of Civil Engineering at TUT in co-operation with Finnish Transport Agency. The ballast material was varied between typical coarse-grained crushed rock aggregate, natural gravel and the gravel partially reinforced with coarse-grained crushed rock. Correspondingly, three of four different sleeper types were prototypes that were chamfered in varying scale from the central section of their bottom surface the fourth being a standard Finnish concrete sleeper BP89. Depending on the test series, cyclic load corresponding to 20 tonnes axle load was repeated until the accumulated loading corresponded to 10–20 MGT. The objective of tests was to estimate the effect of the ballast material on the bending stresses met by the sleeper and, additionally, the effect of simple sleeper model modifications on bending stresses. The strains and vertical displacement of the railway sleeper and strains of the subballast were measured dynamically throughout the test. Based on all the measurements together it was possible to estimate the interaction of the railway sleeper and support layer quite comprehensively. The strain of the top surface of a sleeper was found to be largely determined by the difference between the recoverable displacements measured next to and in the middle of the track. Typically the deflection (bending flexure) of 0.5 mm caused about 150 μm/m strain to the upper surface of the sleeper. Sleeper which had the highest and longest chamfer in the middle section of its bottom surface behaved promisingly considering the weaker support from gravel ballast. Maximum strains in the sleeper increased as the support from ballast decreased.