Blast-Induced Air and Ground Vibrations: A Review of Soft Computing Techniques

Abstract

With favorable fragmentation during blasting, environmental issues such as ground vibration and air overpressure (AOp) have remained challenging issues for any mining or civil engineering project. Explosives’ accessories have been developed over the years such as ordinary electric detonators, millisecond delay detonators, and electronic delay detonators. Geomechanical properties of rock mass, explosives charge per delay, and distance between blast and monitoring point plays an important role in these two environmental issues of blasting. Airblast or AOp when gases are vented out during explosion to the atmosphere through various mechanisms such as rupturing or rock, blowing out of stemming material, displacement, and colliding of rock during blasting. Many researchers developed empirical equations for prediction of ground vibration. Similar equations were developed for prediction of AOp based on maximum charge per delay and distance. Empirical equations or statistical methods were not accurate for prediction of these environmental issues. During last decade, various artificial intelligence and machine learning techniques such as artificial neural network, neuro-fuzzy, fuzzy logic, support vector machine, and various hybrid models were developed with acceptable accuracy for prediction of ground vibration and AOp resulting from blasting. The mentioned models were reviewed and discussed in detail with their used input variables and accuracy and the best models among them were highlighted and suggested to be used by researchers and designers.