A new laboratory facility has been designed to test a 1.2m × O.6m × O.2m prismatic sample under biaxial load. The system will be used for large-scale tests on artificially-made fractured rock models. The initial plans called for a system mounted on wheels to facilitate its assembly before each lest, it was also planned to use a reduced number of hydraulic jacks and to distribute the load evenIy on the sample via a pyramid of simply supported elements. But due to the complexity of the system, the final cost rose to levels far beyond those initially considered. So it was necessary to design a new system adjusted to the economical restrictions. The final design results in a more compact biaxial testing frame, and while it maintains most of the benefits of the first design, it allows simulating other loading conditions such as those encountered in open excavations.
The Strength of rock masses can be estimated by several methods. such as: back calculation of well documented case histories, numerical analysis of jointed rock masses and laboratory tests on models of jointed rock masses. It is difficult to obtain a large assortment of well-documented case histories. Numerical methods can be used to analyze several cases at low cost, but there is a need to evaluate the capability of these methods to simulate the real behavior of jointed rock masses. Numerical methods can be checked against well-documented case histories and against model tests. Model tests, although more time consuming than numerical tests, may shed some light on new failure modes, not yet disclosed by numerical codes. A research program was funded by FONDECYT (The Chilean Foundation for Science and Technology) in order to evaluate the strength of large-scale rock mass models with discontinuous joints. This paper describes the design of the laboratory facility that will be used to carry out the model tests.
Although it would be ideal to perform triaxial tests, the complexity of the loading system is such that it becomes more practical to perform biaxial tests. The sample has to be thin in order to minimize the influence of stresses perpendicular to the loading plane, approaching a situation of plane stresses. On the other hand, the sample must have a certain minimum thickness in order to prevent buckling. The next question is how to load the sample. The following considerations apply:
The loading system should apply a uniform stress on the external races or the sample
The loads should be perpendicular to the face, with no shear stresses at the contact. Contact shear stresses can result in a strength increase of up to 20%.
The loading system must follow the deformations of the sample, it should allow the free deformation of the loading boundary.
The loading system should be as stiff as possible, in order to observe the post peak behavior. During the design stage, an effort was made to satisfy the first three restrictions.
