| In recent years,the construction of highways,urban subways,and mine tunnels have been expanded because of the development of the economy.In the process of tunnel excavation,water diversion structures such as collapse columns and faults are often encountered.Especially in steeply inclined water-rich faults with large dip angles,water inrush accidents are very easy to occur,which seriously threaten the tunnel construction and safety production.In recent decades,water inrush accidents in China have caused more than ten thousand casualties and huge economic losses.However,it is still an urgent problem to accurately predict the collapse column and water-rich fault in front of the tunnel face.The direct current resistivity method has the advantages of a portable device,strong anti-interference ability,and high sensitivity to conductivity water-bearing structure.The resistivity method is used to predict the distance of the low resistivity body in front of the tunnel face based on the minimum apparent resistivity observed in the pole-dipole array.It has been widely used in tunnel advanced detection and mine water disaster prediction and prevention,which has made many achievements.So far,the models of advanced detection in tunnels are based on the resistivity isotropic medium model.Inthe actual process of underground coal mining and underground tunnel driving,the coal seam itself has obvious resistivity anisotropy.Water-rich structures such as collapsecolumns and steeply inclined water-rich faults in front of tunnels often show anisotropy of resistivity due to rock fracture and joint development.Many studies have shown that resistivity anisotropy leads to unexpected bias in data interpretation of the electromagnetic method.However,due to the complexity of three-dimensional numerical simulation of the resistivity anisotropy in the whole space,the resistivity anisotropy of the medium has not been considered in the tunnel advanced detection of the resistivity method,so the anisotropy study of anomaly resistivity in front of tunnel face is helpful to prevent water inrush accidents.In this paper,a three-dimensional unstructured finite element numerical simulation of resistivity anisotropy in whole space has been realized.On this basis,the apparent resistivity profile curves of the pole-dipole array for the anisotropic resistivity cube and low resistivity steep fault in front of the tunnel face are calculated by the conversion coefficient of the potential observed on the pole-dipole array.It is the first time to establish a tunnel prediction model under anisotropic resistivity conditions for the whole space model.Considering the complexity of the actual geological structure,the size of anomalies is variable and the resistivity distribution is arbitrary,the prediction formula obtained from the numerical simulation results or a simple physical model has great uncertainty and its reliability is difficult to evaluate.In this paper,the randomization idea of the Monte Carlo method is used to solve this kind of nondeterministic problem.The resistivity values of three-dimensional unstructured grid elements of the cube abnormal body and steeply inclined water-rich faults are randomly processed under the condition that the resistivity is low compared with the surrounding rock resistivity.Anisotropy coefficients of water-rich faults also vary randomly to simulate arbitrarily complex low-resistivity water-bearing structures in front of the tunnel face.The results of three-dimensional finite element numerical simulation based on 10000 random models show that 94.30%of the prediction models for collapse columns water-rich anomalies have errors within 15%.For the prediction models of vertical fault anomaly,85.36%of the prediction results have errors within 10%.In addition,93.17%of the prediction results have errors within 15%when the prediction equation based on vertical faults was used to simulate random faults with 75 degrees.These results all show good normal distribution characteristics,and the predicted distance of most random anisotropic anomalous bodies is in good agreement with the actual distance from the tunnel face.These errors are small,which verifies the accuracy and reliability of the prediction models in this paper.Monte Carlo method provides a new method for quantitative description of accuracy and reliability of advanced detection in tunnels.In the above traditional direct current resistivity method of advanced detection in tunnels,the point current source electrode is arranged in the position close to the tunnel face,and make the position of the point current source electrode is as close as possible to the abnormal body in front of the tunnel face.The stable current field generated by a point current source excitation is coupled with the low resistivity body to produce the apparent resistivity minimum abnormal response.It is observed by the apparent resistivity profile arranged behind the tunnel face and it is used to predict the position of the low resistivity body in front of the tunnel face.The biggest problem is that the apparent resistivity anomalies observed in the tunnel floor behind the tunnel face are very weak,which results in the small detection depth of advanced detection in tunnels by the traditional direct current resistivity method by a point current source excitation.Although scholars have suggested that changing the excitation current source or changing the combination of transceiver devices can enlarge the detection depth of the anomaly to a certain extent.After all,because of the inaccessibility in front of the tunnel space,the effect of increasing detection depth is limited.In this paper,the horizontal pilot hole drill pipe is used as a long electrode current source excitation by taking advantage of the characteristics of requiring horizontal drilling holes in the tunneling process.Thus,the current field can be excited closer to the anomalous body in front of the tunnel face and coupled to produce a stronger apparent resistivity anomalous signal.Based on the three-dimensional unstructured finite element numerical simulation of whole space long electrode source excitation by direct current resistivity method,theoretical potential and apparent resistivity formulas of whole space long electrode source for advanced detection in tunnels are firstly proposed in this paper.The accuracy and reliability of the proposed three-dimensional unstructured finite element algorithm are verified by theoretical tests on the whole space model of the long electrode source.Under the condition of the traditional tunnel floor observation fault models aresimulated to obtain the apparent resistivity response of the long source electrode,and then extract the minimum position of apparent resistivity.By using the multivariate linear regression fitting method,the prediction model of the long electrode source is put forward for the first time.The R square of the model equation is equal to 0.992,and the errors between 91.6%of the predicted results and the actual fault distance are small,which indicates that the prediction model of the long electrode source is accurate and reliable.Finally,based on three-dimensional finite element numerical simulations of the fault model,the prediction models of long electrode source and point source electrode for advanced detection in tunnels are compared under the same conditions,the results show that the prediction model of point electrode source only can accurately estimate the position of the shallow anomaly,but the prediction model of long electrode source for shallow and deep abnormal position estimation is very accurate.The long electrode source has a significant advantage over the point electrode source in detecting depth.The prediction model of the long electrode source provides a quantitative method to estimate the position of the water-bearing anomaly in front of the tunnel face.This long electrode method increases the detection depth of advanced detection in tunnels,which is of great significance to improve construction efficiency and guarantee tuunel safety. |
PDF Full Text Request