A novel experimental technique of split Hopkinson bar system is introduced to improve the incident wave against wave dispersion. By this method, the incident loading wave is of less oscillations and no rising time, which is very essential to obtain accurate data from dynamic tests on rock-like materials by using the Hopkinson bar test system. From the dynamic test results on Bukit Timah granite at high strain-rate up to 103/S, it is found that the dynamic mechanical property of granite is much sensitive to the loading conditions.
The split Hopkinson pressure bar (SHPB), first introduced by Kolsky (1949), is usually used to measure dynamic properties of material within the strain-rate range of 10 2/s to 10 4/s. Investigations on dynamic properties of rock like materials have become interesting recently on the fracture behaviours (lambert 2000; Zhang et al. 2000), compression properties (Tedesco 1998; Shan 2000) or tensional properties (Eibl 1999) at high strain-rates by the SHPB system. To the rock-like materials, a large diameter Hopkinson bar system is needed to ensure the representation of the specimen, and thus reducing an even strong wave dispersion effect during the tests. Because of the more brittleness or the very small damage strains of Such rock-like materials, it cannot get accurate test results by the traditional SHPB technique. So, in order to make it possible to get correct dynamic data form the SHPB system for such materials, it is essential to solve the influences of elastic wave dispersion on the test accuracy. Many researches have been performed in such area, such as numerical method (Gong et al. 1990, Zhao 1998) or experimental method (Li et al. 2000). Here, a new experimental method that is much simpler and more effective is introduced. With this method, five series of dynamic tests on the Bukit Timah granite were conducted at the compressive strain-rates of 46/s, 101/s, 186/s, 335/s and 874/s. From the dynamic stress-strain data, it was found that the mechanical properties of granite are more sensitive against the strain-rate.
In the compression set-up of the Hopkinson bar apparatus (as illustrated in Figure l), two long uniform cylindrical steel pressure bars, called incident bar and transmitter bar, are supported coaxially. A short cylindrical specimen is sandwiched between these two bars. According to the one dimensional elastic wave theory, when the cylindrical striker bar of length L impacts the cylindrical incident bar at velocity V (V can be detected by the parallel illuminant.), compressive waves are produced at the interface of the two bars, and propagate at the constant velocity of Co (equal to (E/ p)1/2 according to the one-dimensional wave theory) into the striker bar and the incident bar in different directions. As the compressive wave reaches the other end of the striker bar, it reflects back and changes into tension type. The particle velocity U after the front of the compressive wave in the steel bar, can be calculated from the conservation of momentum equation.
