Comparative laboratory testing has been undertaken on samples of Hawkesbury Sandstone in order to evaluate the relative importance of currently recommended procedures for specimen preparation and testing of core samples of weak rocks. The test results indicate that many of the recommendations relating to the measurement of unconfined strength, Young's modulus and Brazilian tensile strength are unnecessarily stringent.
Des tests comparatifs ont ete enterpris en laboratoire sur des echantillons de grès de la region du Hawkesbury afin de pouvoir evaluer l'importance relative des procedes couramment employes pour la preparation et l'etude des echantillons de roches à structure fragile. Le resultat des analyses indique que nombre de recommandations concernant l'evaluation d'une force unidirectionelle, soit le Module de Young et le test Bresilien d'elasticite d'un echantillon, sont d'une severite trop rigoureuse.
Laborvergleichsteste sind an Proben von Hawkesbury Sandstein vorgenommen worden, um die relative Bedeutung der zur Zeit vorhandenen Verfahren zur Probeherstellung und Untersuchung von Kernproben schwachen Gesteins zu bewerten. Die Untersuchungsresultate ergaben, daß viele Angaben bezueglich der Messung der einachsigen Druckfestigkeit, Young's Modulus und indirekter Zugfestigkeit unnötig streng gefaßt sind.
Standards for the preparation and testing of rock core samples for unconfined compression, triaxial and Brazilian tests have been published by many authorities such as the International Society for Rock Mechanics (ISRM, 1972), the Canada Centre for Mineral & Energy Technology (CANMET, 1977), the American Society for Testing & Materials (ASTM, 1974) and the Standards Association of Australia (SAA, 1981). Many of the recommendations given by these and other similar authorities clearly come from common sources. For example the Draft Australian Standards are largely based on the ISRM recommendations which in turn reflect the efforts of Bieniawski and Franklin, based on work at the CSIR in Pretoria and Imperial College but also on comparative test data from the US Bureau of Mines (Hoskins & Horino, 1968). The Canadian recommendations are largely based on the ASTM Standards which in turn are also strongly influenced by the work of the US Bureau of Mines. It is clear that the sample preparation and test procedure recommendations for compression and tensile-tests on rock core arose from research organizations largely concerned with mining rock mechanics and, in particular, hard rock mining. Thus most of the rock types used for the comparative test studies had unconfined compressive strengths greater than 50 MPa; in fact the majority greater than 100 MPa. Now in terms of civil engineering rock mechanics, particularly foundation engineering, the majority of rocks of concern have substance strengths of less than 50 MPa. For example in the Australian context, the Triassic sandstones and shales of the Sydney Basin have unconfined strengths generally in the range 5 to 25 MPa while the Silurian mudstone beneath Melbourne ranges in strength from about 1 MPa to 10 MPa. In testing such weak rocks the question arises as to the applicability of the specimen preparation and test procedure recommendations given by the various authorities discussed above. In order to investigate this question a programme of testing was undertaken using samples drilled from three blocks of Triassic Hawkesbury Sandstone taken from adjacent to one another in the summers by quarry north of Sydney. By testing this sandstone in both the oven dry and saturated states it was possible to carry out comparative testing on materials with unconfined strengths of 50 MPa and 20 MPa.
The specimen preparation parameters and test procedures that were investigated were all related to the recommendations set out by the ISRM because these recommendations have formed the basis for various national standards. All testing was carried out in an Instron servocontrolled test machine and all comparisons are made against a data base of samples prepared and tested fully according to the ISRM recommendations. Table 1 lists the number of specimens tested to investigate each of the preparation or test procedure aspects listed above. All specimens were nominally 53mm diameter. Considering Table 1 it can be seen that a statistically meaningful number of tests was performed for the uniaxial test and Brazilian test comparisons. However, the specimen numbers for the Young's modulus tests are insufficient for statistical evaluation and thus only gross influences are noted. Determination of Brazilian Tensile strength effect of testing directly between test machine platens (i.e. no test jig) effect of using cardboard strips against test machine platens (again no test jig) Space does not permit a detailed explanation of the test procedures adopted in this programme. However, the following notes give some indication of what is meant by the brief description given in Table 1. Great care was taken in ensuring that the 'saturated' specimens were at a consistent degree of saturation. This was achieved by repeated application and release of vacuum to the specimens stored under water.