A New Design Method For Drilled Shafts In Rock ? Synopsis

The design of side resistance of drilled shafts socketed into rock has traditionally been empirically based on the results of pile loading tests in which the side resistance can be isolated from the side resistance. In practice, this is achieved in one of several ways : loading the pile in tension, using a collapsible base, or by using a device such as an Osterberg cell to load the pile from the base. Although this approach has resulted in generally safe design practice, this has been at the expense of efficient design practice. Safety in design has been achieved through conservatism, which has been necessary because of the large scatter that is evident in the pile socket resistance database (Rowe and Armitage, 1984; Kulhawy and Phoon, 1993). Research at Monash University through the 1980's has been aimed towards understanding and then modelling the complex mechanisms of shear transfer at the concrete-socket interface of sedimentary rocks. Armed with such an understanding, the important factors which affect the strength of pile sockets can be isolated, and the reasons for the large scatter can be deduced. This paper briefly describes the research approach taken, and the main features of the design method and computer program which have now been developed. Some load-settlement predictions of the program are compared with the results of full-scale pile socket tests. The paper then presents the results of a numerical study which relates predicted socket a factors to unconfined compressive strength of the rock, socket diameter, initial normal stress and socket roughness.