Study On Processing And Evolution Of Granite Vibration Signals Under Cyclic Blasting

Blasting technology is often used in engineering fields such as tunnel excavation,mining,and construction of water conservancy and hydropower facilities.It can significantly speed up the construction progress of the project and improve the operation efficiency.The social and economic benefits brought by it are very obvious.Because the blasting construction against the rock mass is a destructive construction behavior,the blasting vibration waves,air shock waves and flying stones generated by it seriously threaten the personal safety of the workers and the structural safety of nearby structures.Especially the blasting vibration effect,which inevitably causes certain damage to the rock itself,has been the focus of experts and scholars.Therefore,It is meaningful to study rock mass blasting vibration effect and control.The research contents and main conclusions drawn in this article are as follows:First,the time-frequency analysis of two typical acceleration zero drift signals in the granite cyclic blasting experiment is carried out,and three main zero drift correction methods are used to perform zero drift correction processing.The analysis results show that: the choice of the time period for processing the signal has a significant effect on the least squares zero drift correction results under the polynomial function and power exponential function,but has no effect on the zero drift correction results of the wavelet function method and the moving average method.As the level increases,the zero-drift component of the single-layer signal decomposed by wavelet increases,and the amplitude range of each layer's signal decreases accordingly.Under different sliding orders,the baseline of the speed signal after the zero-drift correction by the moving average method is corrected.As the number of slip points increases,it shows a downward trend.By comparing the degree to which the baseline of the speed signal after the correction deviates from the zero point baseline,the number of slip points that makes the zero drift correction result optimal for each slip order can be found.Next,in view of the shortcomings of the first three zero drift correction methods,we find a set of empirical mode decomposition(EEMD)algorithms that can be adaptively decomposed,and the effect of the white noise coefficient of the EEMD algorithm on its zero drift correction result is mainly studied.The results show that the EEMD method combined with high and low frequency processing can remove the zero drift trend of the blasting acceleration signal.As the white noise coefficient increases,the correction index increases in different frequency bands to varying degrees,and thetwo show power exponential relationship of a high correlation.By modifying the index and combining the frequency response range analysis of the sensor,the optimal value range of the white noise coefficient corresponding to the zero drift signal is determined.The optimized noise figure makes the EEMD method minimize the signal's zero drift tendency while preserving the spectrum information of the signal as much as possible.Finally,the zero drift correction results are used to perform linear regression analysis on the vibration signal,and the main frequency evolution of acceleration signals at different measurement points is analyzed.The analysis results show that,compared with the peak acceleration of the particle,the Sadofsky's formula is more suitable for describing the law of the attenuation of the peak velocity of the particle under the same conditions.During the cyclic blasting,the main frequency of the acceleration signal will fluctuate locally,rather than decay steadily with the increase of the blast center distance.At the same time,based on the Holmquist-Johnson-Cook(HJC)constitutive model,numerical simulation of granite cyclic blasting is implemented in ANSYS/LS-DYNA software.By analyzing the similarities and differences between the numerical simulation signal and the zero drift correction signal,the rationality of the correction result is confirmed on the other hand.