Evaluation of the block punch index test with particular reference to the size effect, failure mechanism and its effectiveness in predicting rock strength

Abstract

Measurement of rock strength requires testing which must be carried out on test specimens of particular sizes in order to fulfill testing standards. Often, the coring process breaks up the weaker core pieces, and they are too small to be used in either conventional strength tests or index tests such as point load test. The block punch index (BPI) test, which requires flat disc specimens without special treatment has been developed during the last decade. This paper presents the results of a study of BPI test device to provide new contributions to previous works on size effect in BPI test using a wide range of rock types, in addition, to assess the effectiveness of the test in predicting rock strength by an experimental way. About 2600 disc specimens were tested, and the results were combined with those from previous works and analyzed using statistical and graphical methods. Finite element modeling was also conducted to evaluate the stress distribution created within the rock specimen by punching action of the test device. The test results suggested that size correction in the BPI test was indispensable. Using a very large test database, the size-correction factors suggested by previous workers were modified in the present study. The finite element modeling indicated that the failure surface initiated at the top of the specimen and propagated into the specimen. This demonstrated that failure was not produced by a true shearing. It was also noted that cohesion values predicted from BPI tests were greater than those obtained from conventional tests. Both of these results suggested that the BPI test was not an accurate device for directly determining shear strength of the rock specimen and should only be used as a strength index. The experimental results showed that the corrected BPI values led to lower errors in determining uniaxial compressive and tensile strength when compared to those from point load testing. Possible uses of the BPI in rock mass classification were also briefly reviewed and a strength classification based on BPI was suggested.