Rock Breakage by Hydraulic Hammer: Comparative Study Between Scale Model in Laboratory and in Numerical Simulations

The dynamic breakage process of rocks by the impact of a hydraulic hammer is more complex than one normally assumes. It is even more difficult when one has to conduct the breakage under water. In such cases one does not know exactly what the surface topography and composition are at the bottom of the sea or river and, hence, it is more difficult to select the best position of the hammer before starting the hammering. Typically, an industrial hammer has a diameter of about 250 mm, a length of about 1.25 to 1.75 m and a mass of several tons. The hammer frequency can be up to 50 impacts per minute. The general principles of rock hammering can be easily summarized, but this knowledge is not sufficient to adapt adequately the method for less common rock conditions (e.g. less brittle rock, presence of vertical and subvertical fractures, etc.). That is the reason why we initiated a research project, whereby a scale model has been developed to conduct experiments in the laboratory and whereby discrete element simulations are being conducted. Both aim at a better understanding of the basic mechanisms of rock hammering. The scaled experimental model allows varying the impact energy and the static weight independently. Within the experiment, the hammer is in contact with the rock specimen under a certain confining pressure and with a certain contact pressure due to the static weight. A weight falls on the hammer, creating the dynamic action or impact; the height over which the weight falls can be varied. The penetration is recorded after each impact. After a certain penetration the test is stopped and the internal crack patterns are described. Some simulations have been conducted, allowing the modeling of individual fracture initiation and growth during the penetration of the hammer by either a constant displacement rate (using the UDEC-code) or by introducing the dynamic process (using the PFC-code). Some of the observed features in the laboratory experiments can be visualized in the simulations. However, further research is needed to allow the simulations being an adequate tool to well understand the dynamic process in laboratory experiments, as well as the in-situ behavior.