Vulnerability assessment of building structures due to underground blasts using ANN and non-linear dynamic analysis

Abstract

Building structures near the mines area are prone to blast-induced ground vibration due to underground blasting. The level of ground vibration typically quantified as Peak Particle Velocity (PPV) is used to assess the vulnerability of the buildings. However, the PPV does not take into account the dynamic characteristics of the structure. To incorporate structural response in the design and assessment process, Artificial Neural Network (ANN) was used on underground blast data from the literature to obtain the Peak Ground Acceleration (PGA). Charge weight (Q), distance from the blasting point to the monitoring point (D) and soil's elastic constant (E) were taken as input to ANN and peak ground acceleration (PGA) was the output. Safe site distance for constructing structures was determined using non-linear time history analysis of structures having periods ranging from short to long duration, incorporating soil flexibility. The main findings highlight the shortcomings of considering PPV as the sole criteria for vulnerability assessment. The structural response such as interstorey drift, and damage level and locations must be taken into account.