Cyclic Behaviors of Railroad Ballast within the Parallel Gradation Scaling Framework

Because of the large grainsizes typical of railroad ballast, large triaxial samples are required to assess the reactions of these materials. The parallel gradation modeling technique was originally developed by John Lowe in 1964 to allow assessment of large grain-size geomaterial properties in smaller, more typical testing facilities. Emphasis has focused on monotonic loading, in which the material is progressively loaded to failure. Cyclic testing of this model has been absent. This paper presents an investigation of the possibility of using the parallel gradation modeling technique in a cyclic triaxial testing framework. Three separate gradations of ballast material were used in this research. The largest gradation contains a top particle size of 63.5 mm (2.5 in.) and is marketed as #3 modified railroad ballast. The second two gradations contained a top size of 38 mm (1.5 in.) and 19 mm ($3/4\mathrm{in}$.), respectively. Up to 10,000 load cycles were applied for each test. Resilient modulus, permanent axial, volumetric strain, and particle shape were determined from the test results. It is concluded that applying parallel gradation technique to cyclic behavior characterization should be cautious. If particle shape is not consistent throughout the particle sizes used in the parallel gradation model, the model is invalid in the cyclic triaxial framework.