Evaluation of a blast damage model based on stress attenuation criteria

Abstract

Empirical models still maintain a large presence in blast design practices despite developments in progressively complex modelling techniques to describe blast damage. This can be attributed to the heterogeneity of rock masses and the related difficulties associated with obtaining the accuracy level required of higher order models within the practical constraints of time and cost. Although empirical blast damage models may have the tendency to misrepresent the importance of critical defining parameters leading to less than accurate outcomes. Where inappropriately sensitive to certain properties, empirical blast damage models can result in less than optimal blast design parameters. This research project aimed to evaluate a new empirical blast damage model developed by Onederra (2016) that has been designated as the iDamage model. This model utilises a stress attenuation function and a tensile strength limiting criteria to yield an estimate for radius of fracturing induced by a single charge. The evaluation consisted of making comparisons to existing empirical models, in terms of damage output as well as in the application of these models to determine estimates for fundamental blast design properties. It was determined that the iDamage model appropriated the changes in the rock mass and explosive characteristics in a more stable capacity, yielding a more central estimate of fracture radii relative to other empirical models. The scenario for a 120mm charge radius in fully confined conditions, for an emulsion product (VOD = 5500m/s) in hard rock is a good representation of this observation. The iDamage model projected a fracture radius of 2.9m while other empirical model estimates gave a range of values between 1.9m and 4.0mThis was further supported by comparison to the damage limits determined from a range of practical case studies. Where the average deviation from the practical scenario was 0.35m for the iDamage models but was as high as 1.75m for other empirical methods. This tendency was also observed when using the damage limits of the iDamage model to determine estimates for blast design parameters. An initial sensitivity analysis generally demonstrated high sensitivity of the iDamage model to parameters that described both the rock mass and explosive (such as σT and VOD). However, for lower yield explosives, explosive characteristics were relatively more sensitive. Based on the often balanced sensitivity to input parameters and relative centrality of damage estimates the iDamage model was proposed to be an effective tool to supplement the initial processes involved in blast design.