1. Introduction
Drilling and blasting method is still an economical and viable method
for rock excavation and displacement in mining industry. The negative
effects of blasting are unavoidable and cannot be completely eliminated
but certainly minimize up to permissible level to avoid damage to the
surrounding environment with the existing structure. Among all negative
effects, ground vibration is major concern to the planners, designers
and environmentalists. In the certain level of strength and frequency,
the ground vibration can sometimes lead to damage of adjacent structure
and may disturb the human comfort and health in the nearby blasting
area. Hence, when mining operation is done near some important civil
structure, prediction and control of ground vibration become critical. |
2. Investigation
An observation of blasting induced ground vibration has been conducted
for safe coal mine activities near a residential area in Sangatta,
Indonesia. A number of researchers have suggested various methods to
minimize the ground vibration level during the blasting. Ground
vibration is directly related to the quantity of explosive used and
distance between blast face to monitoring points as well as geological
and geotechnical conditions of the rock units in the working area.
Geological and geotechnical conditions and distance blast face to
monitoring point cannot be altered but the only factor, i.e. quantity
of explosive can be estimated based on certain empirical formulae
proposed by the different researchers to produce ground vibrations in a
permissible limit. In the present research, few important and widely
used predictors, such as USBM equation, Ambresseys-Hendron equation and
Langefors-Kihlstrom equation, have been used to predict the peak
particle velocity (PPV) and computed results are compared with actual
field data. The same input-output data sets have been also used for the
prediction by artificial neural network (ANN). The basic idea is to
find the scope and suitability of the ANN for prediction of PPV over
the widely used vibration predictors. |
3. Result
Based on the study, it is established that the feed-forward
back-propagation neural network approach seems to be the alternative
option and appropriate prediction of PPV to protect surrounding
environment and structure beside the predictor equations. The predicted
value of PPV from the neural network is closer to the recorded PPV than
the one predicted using the predictor equations. It can be inferred
that the predictor equations which are developed by attenuation
relationship equation are not the only acceptable approach that can be
taken into account in predicting the PPV. |
|