Soil suction response of granular railway formation materials under cyclic loading

Abstract

Changing climatic conditions worldwide are causing a change in moisture conditions of railway formations and slopes, thereby either strengthening or weakening them. Current railway formation design methods do not take into account the changing moisture conditions over the predicted life of a railway formation. These changes in soil strength are due to the influence of soil suctions. The basic principles of unsaturated soil mechanics have been well established in the field of geotechnics, and this study joins an international body of work that seeks to apply unsaturated soil theory to the field of railway and pavement formation materials. This study comprised of box testing of subgrade and subballast formation materials at different moisture contents. These formation models were instrumented with tensiometers to monitor the effect of cyclic loading on soil suctions under typical heavy haul loading magnitudes (26 tonne/axle). The materials were tested under a number of different loading frequencies subjected to the same cyclic heavy haul load magnitude. The study investigated the suctions present in railway formation materials, as well as the application of suction instrumentation in railway conditions. Initial testing of the tensiometers showed evidence of interference caused by the stress imposed by the soil skeleton on the tensiometer housing. This prompted a development program that resulted in an improved tensiometer model that was isolated from the effects of external loading and could therefore successfully measure the suctions present in the formation material. Suctions were successfully measured in both the subgrade and subballast material under various moisture contents demonstrating the different suction magnitudes generated by the materials under expected formation conditions. The subballast material was found to generate suctions of between 1 and 15 kPa over a wide range of degrees of saturation (Sr: 0.32 – 0.96), while the subgrade material was found to generate a greater range of suctions between 1 and 95 kPa with a smaller variation in degree of saturation (Sr: 0.61 – 0.88). The deformability of both the subgrade and subballast materials were affected to different degrees as a function of the soil suctions. The subgrade material was found to be dependent on the moisture state and soil suctions to a greater degree than the subballast material. A deformation of 141 % of the failure criterion was observed at low suctions (1 – 5 kPa) in the subgrade material at loads similar to those experienced by South African heavy haul railway formations. In contrast, the deformation significantly reduced to 55 % of the failure criterion when the suctions increased by a relatively small amount ( 10 kPa). Further desaturation of the subgrade material resulted in higher suctions (40 – 95 kPa) strengthening the formation further to a final deformation of 47 % of the failure criterion. It was found that subballast deformation was significantly affected by the loading frequency and less by moisture state. Due to the low suctions present in the subballast material, the maximum deformation observed was 61 % of the failure criterion under effectively saturated conditions with suctions B1 kPa. Under the highest suctions present in the subballast material ( 14 kPa), the material only strengthened to the point where 47 % of the failure criterion was observed. The application of geotechnical testing methods to granular formation materials are also investigated and discussed, in terms of the effectiveness of tensiometers and filter paper methods in determining suction, as well as considerations when sampling granular soil for moisture content. The study therefore provides important information for laboratory testing of formation materials, the techniques required to successfully measure suctions in railway formations and the implications of cyclic loading on formations at various degrees of saturation.