PERFORMANCE OF LARGE-SCALE MODEL SINGLE TIE-BALLAST SYSTEMS

Large-scale models of a single tie-ballast system were constructed over artificial ballast support that had variable compressibility ranging from rigid to very flexible (California bearing ratio = 10). Test configurations included a 0.45-m depth of crushed limestone ballast conforming to an American Railway Engineering Association grading No. 4. A steel footing 920 mm long by 250 mm wide by 150 mm deep was used to model the bearing area of a typical tie below the rail seat (i.e., one rail). Each rail seat was subjected to a repeated load of between 20 and 150 kN for a (typical) duration equivalent to 12 million gross tonnes in track. The principal objectives of the experimental work were to investigate the influence of load level and ballast support compressibility on the rate of accumulation of permanent deformations and ballast degradation. The test results show that at a given load level the rate of tie settlement is quite sensitive to ballast support compressibility. A competent ballast support resulted in a deformation-log tonnage response that was essentially linear. However, progressively weaker supports gave increasing semilogarithmic rates of settlement with tonnage. For a given support compressibility, a critical load level was identified that, if exceeded, led to a dramatic increase in settlement rate. The critical load level was also identified as a threshold level above which the generation of fines in the ballast directly below the tie was observed to increase markedly.