INTRODUCTION
Predicting ground vibrations (and noise) induced by blasting is becoming increasingly important. This is due in large part to the fact that more rock excavation is going on near populated areas, e.g. due to urban sprawl encroaching on quarry and mine operations, and in part to an increased sensitivity to environmental disturbances (du Pont, 1977, p. 423). In response, it is likely that more stringent and more complicated limits will be imposed on blasting operations (e.g. Office of Surface Mining, 1983). Several methods exist for predicting ground movements induced by blasting. Such methods must be relatively simple in order to be acceptable in common engineering practice. This largely precludes the use of numerical simulations. Nevertheless, it must be recognized that ground vibrations induced by blasting are the consequence of a complex series of events. Presently used predictors cannot account for all these, and produce large deviations of the actual values from the predicted values. Although it is unlikely that all of these can be accounted for in detail by means of simple closed-form expressions, it is probable that better predictability can be achieved by accounting explicitly for dominant aspects of pulse initiation and propagation. The method proposed here recognizes that site-specific predictors are most likely to be successful. The empirical parameters of the model proposed here, ideally, also are to be determined at the location of concern. The general form of the model is derived by incorporating separately the descriptors of geometrical and of inelastic attenuation.
BLAST VIBRATION PREDICTORS
The generation and propagation of ground vibration from blasting has gained the attention of seismologists and mining engineers for a long time. Early Bureau of Mines works (Blair and Duvall, 1954; Duvall and Petkof, 1959) tried to correlate the amplitude of the seismic pulse with the amount of charge and with the distance from the source. Assuming spherical symmetry, the conclusion was that any linear dimension should be scaled to the cube-root of the charge size. Therefore, measurements of different blasts with different charge sizes taken at various distances should be scaled to the equivalent distance, called a scaled distance, which is the actual distance divided by the cube-root of charge weight. Similar results were obtained by later investigators (Ambraseys and Hendron, 1968; Dowding, 1971). The amplitude of seismic waves and scaled distance were related by the inverse power law: (mathematical equation)(available in full paper) Here, V = particle velocity, D = distance, W = maximum charge per delay, and K and n are empirical constants derived from the best-fit straight line of V vs. D/W 1/3 on a log-log plot. If long, cylindrical charges are used, the explosive geometry can be assumed to be cylindrical rather than spherical. From dimensional analysis, it can be concluded that any linear dimension should then scale with the square root of the charge weight (Devine, 1962; Devine and Duvall, 1963). Blasts should be scaled to the equivalent distance or scaled distance, defined as the actual distance divided by the square root of the charge weight.