Department of Geology, University of New Brunswick, Fredericton, New Brunswick, Canada

The tensile strength, and its variation with direction, was determined by the pointloading and the line-loading (Brazilian test) methods for nine rocks from five sandstone formations in New Brunswick, Canada. Altogether 319 double point-load and 357 line-load tests were conducted on discs having a diameter of 22.9 mm and thickness of 7.9 mm.

The tensile strength determined by the point-load test and as calculated by the Frocht formula was found to be consistently lower than that obtained by the line-loading technique. To some extent, the difference is due to the fact that most of the sandstones are anisotropic with respect to tensile strength and that the constraints in the point-load test are such that fracture occurs along the plane of minimum tensile strength; the computed tensile strength therefore represents a minimum value. To a greater extent, the lower tensile strength calculated from the point-load test is due to the difference in the states of stress induced by the two techniques. When correction is made for the anisotropy, the tensile strength from the Brazilian test is still 3.76 times larger than the one obtained from the point-load test.

The Brazilian test seems to yield a more accurate definition of both the tensile strength and its variation with direction. The point-load test is more suitable for the determination of the minimum value of tensile strength and the direction of the preferred fracture plane.