The split Hopkinson pressure bar (SHPB) system is an important tool for determining the properties of rocks at high strain rates. To better appreciate the dynamic behaviour of rocks, it is necessary to understand the limitations of a SHPB and the resulting test data. These details arc discussed in the paper. A 75 mm diameter SHPB system for testing rocks and other brittle materials has been fabricated. The design of this 75 mm diameter SHPB is a significant departure from a conventional SHPB because the input and output bar lengths are considerably shorter. To compensate for the shorter lengths, a novel impactor is introduced. In this way, many of the attendant problems such as wave oscillation and non-uniform specimen stress (or strain) state are eliminated. Details of this SHPB and some dynamic test results of rocks are presented to reinforce the validity of the system.
In 1914, Hopkinson proposed a long elastic bar to study the pressures produced by the impact of a bullet and by the detonation of an explosive device. He recognised that as long asthe pressure bar remains elastic, the displacements in the Pressure bar are directly related to the stresses, and that the wavelength in the bar is related to the duration of the impact obtained from the velocity of sound travelling in the bar. Subsequently, Kolsky(1949) introduced a revolutionary split Hopkinson pressure bar system. He demonstrated how the dynamic stress and strain within the deforming specimen are related to displacements in the elastic pressure bars. The split Hopkinson pressure bar (SHPB) is used extensively for determining the dynamic behaviour of non-metallic and non-homogeneous materials such as concrete and rocks (Kumar, 1968; Li & Gu, 1994; Ross et al 1995; Zheng, et al 1999). However, users of such equipment must take note of the following fundamental assumptions:
wave propagation in the bars can be described by one dimensional wave propagation theory;
stress and strain in the specimen are uniform in its axial direction; and
specimen inertia and friction effects are negligible. Although assumptions (a) and (b) can be achieved for small diameter SHPB tests on metals, Davies and Hunter(1963) and Follansbee and Frantz(1983) showed that high frequency dispersion overriding the main wave is inevitable. Further, Bertholf (1974) suggests that an optimal length/diameter ratio of the specimen should be chosen and that a highly polished specimen/bar interface condition can minimise both inertia andfriction effects. However, when such tests are conducted on non-metal heterogeneous brittle materials such as rocks, a largerdiameter bar is required. A large diameter SHPB system necessitates a correspondingly longer length of the bars. Consequently, the applied strain rate on the material is limited due to an increasein themass. In long bars, the conventional rectangular wave loading that result from the impact will cause strong wave dispersion and oscillation. Further, the assumption of one dimensionalstress and uniform stress/strain in the specimen is no longer valid due to the short duration loading.