MODES OF FAILURE AND REMEDIATION TECHNIQUES FOR ROCK CUTS IN SEDIMENTARY STRATA SUBJECT TO DIFFERENTIAL WEATHERING

Rocks subject to differential weathering usually occur as sedimentary sequences in which harder strata (siltstones, sandstones, and limestones) alternate with softer strata (shales, claystones, and mudstones). Such rocks are jointly referred to as the mudrocks and they dominate the geology in the adjacent states of Ohio, Pennsylvania, and West Virginia. When new road cuts are made through this type of stratigraphy, softer rock layers weather faster than the harder layers. The differential rate of weathering leads to undercutting of the harder layers. Undercutting promotes a variety of slope movements such as rockfalls, plane failures, and wedge failures that could not occur if the support below the competent layers is not removed by the process of differential weathering. Furthermore, the timing of these failures is difficult to predict. There are numerous examples of such failures along the roadways in Ohio, Pennsylvania, and West Virginia some of which are presented in the full paper. Because of the high speed, suddenness of occurrence, and sometime large volume of rock involved, undercutting-induced failures can be quite hazardous.

A number of remedial measures are currently being used by various states to reduce the potential for undercutting-induced failures. These remedial measures can be divided into two categories: (1) those that retard the rate of undercutting or prolong the initiation of undercutting, and (2) those that minimize the potential for occurrence of undercutting-induced failures and the associated hazard. The paper provides an overview of both types of remediation techniques. For a timely implementation of these remedial measures, it is essential to estimate the anticipated rate of undercutting for a given site. A methodology for predicting the rate of undercutting using the second-cycle slake durability index (Id2) of the softer layer is described. The amount of undercutting was measured for 14 road-cut sites in Ohio, Pennsylvania, and West Virginia. The maximum amount of undercutting at each site was divided by the age of the cut to obtain the rate of undercutting. The rate of undercutting was correlated with Id2 values of the respective undercutting units. It was found that the rate of undercutting can be predicted with a reasonable degree of accuracy from the Id2 value of the undercutting layer.