考虑自由面参数影响的地铁隧道爆破振速预测公式优化研究

Optimization of prediction formula for blasting vibration velocity of subway tunnels considering effects of free surface parameters

预测地铁隧道爆破施工产生的振动速度,大多采用萨道夫斯基公式进行计算,很难反映岩体爆破中自由面的影响。然而,在隧道爆破过程中,岩石的自由面面积及自由面数量都会影响爆破地震波的传播,导致振速发生变化。为进一步研究自由面对爆破振速的影响,在萨道夫斯基公式的基础上,引入自由表面积S_(r)、自由面数量系数k和自由表面指数β来表征自由面对爆破振速的影响,并采用统计学t分布进行优化,建立隧道爆破地表振速改进预测公式。以重庆地铁18号线富歇区间隧道工程为依托,开展爆破振动测试,将改进公式与原公式对实测振速的拟合结果对比分析,验证了改进公式的正确性。通过数值模拟建立了3种自由面参数不同工况下的隧道爆破模型,根据改进公式预测结果的各项误差评价指标,可知和方差由3.19×10^(-5)下降为1.29×10^(-5),均方根误差由1.56×10^(-3)下降为9.96×10^(-4),决定系数由0.887上升为0.956,调整决定系数由0.882上升为0.954,证明了改进公式具有更高的预测精度。研究成果可为类似隧道爆破工程提供一定的理论依据。

Predicting vibration velocity generated by subway tunnel blasting construction is mostly calculated using Sadovski formula,it is difficult to reflect effects of free surface in rock body blasting.However,in process of tunnel blasting,area and number of free surfaces of rock can affect propagation of blasting seismic wave to cause change of vibration velocity.Here,to further study effects of free surface on blasting vibration velocity,based on Sadovski formula,free surface area S_r,free surface number coefficient k and free surface index β were introduced to characterize effects of free surface on blasting vibration velocity,and the statistical t-distribution was used to do optimization,and establish an improved formula for predicting tunnel blasting surface vibration velocity.Based on Fuxie interval tunnel project of Chongqing Metro Line 18,blasting vibration testing was conducted.The measured vibration velocity fitting results using the improved formula and the original formula,respectively were analyzed contrastively to verify the correctness of the improved formula.3 tunnel blasting models were established through numerical simulation under different working conditions with different free surface parameters.According to various error evaluation indexes for the predicting results with the improved formula,it was shown that the sum variance SSE(sum of squares for error) drops from 3.19×10~(-5) to 1.29×10~(-5),the RMS(root mean square) error RMSE(root mean square error) drops from 1.56×10~(-3) to 9.96×10~(-4),the determination coefficient R-square increases from 0.887 to 0.956,and the adjusted R-square increases from 0.882 to 0.954 to reveal the improved formula having higher prediction accuracy;the study results can provide a certain theoretical basis for similar tunnel basting projects.