Impact of sample selection on unconfined compressive laboratory test results in hydrothermally veined rocks

ABSTRACT: Geotechnical properties measured by unconfined compressive strength (UCS) laboratory tests can be greatly influenced by sample selection practices, especially for heterogeneous, hydrothermally veined and altered rocks. This data is not captured by many site investigation programs since it is common practice to test homogeneous specimens but discard heterogeneous specimens. This may lead to erroneous assessments of rockmass strength and behaviour prediction for excavations. This paper first discusses industry sample selection practices and a case study of sample selection from the Bruce Deep Geological Repository project. Secondly, the effects of sample selection on UCS test properties of 39 UCS specimens from the Legacy skarn ore deposit are compared using four methods of sample selection for homogeneous and heterogeneous (veined) granodiorite and calcareous mudstone units. In all units there are significant differences in UCS properties between homogeneous and heterogeneous samples. Comparing the total UCS test suite to mixed subsets of each unit shows fairly similar results, but this may depend on how the subset selection method is implemented.