Study On Ultimate Bearing Capacity Of Anchor Bolts Based On Improved Grey Neural Network Model

The anchor and anchoring technology in the field of geotechnical safety is an important branch of the geotechnical industry.With the development of our country's economy and the vigorous promotion of geotechnical engineering,anchor bolts have been more and more widely used in the field of rock and soil anchorage,such as building foundation pits,mine roadways,highway slopes and underground caverns and many other projects.According to the differences in the internal structure of the bolt and the cement mortar and the bond of the anchor,the cement grouting bolt can be divided into the following types: tensile bolt,pressure bolt and shear bolt.At present,most of the projects are using tension type anchors.The ultimate bearing capacity of anchor bolts is not only affected by the strength of raw materials and cross-sectional area of anchor steel bars,but also depends on the anchoring quality of cement mortar.The primary factor affecting the ultimate bearing capacity of anchor bolts is the anchoring quality of cement mortar.However,the complex internal structure of rock and soil and the changing environment of the outside world,leading to the pull-out bearing capacity of the bolt has always been a difficult point in the project.This article is aimed at predicting the important bearing capacity of the anchor rod,mainly from the pull-out test and electromagnetic lossless non-destructive bearing capacity of the two aspects of the following research:(1)With regard to the influence of ideal anchor rods and anchor bolts of different qualities on the ultimate bearing capacity,the drawing results of the anchor bolts are plotted with displacement P-load S curves,and the reasons are analyzed.(2)Based on the original GM(1,1)model,based on the P-S curve of the anchor under different conditions,it is proposed to use a hyperbolic curve fitting,and then calculate the adaptive weight of the grey model based on the fitted residual.The modified GM(1,1)model is used to fit the anchor P-S pattern,and finally the ultimate bearing capacity of the bolt is predicted.(3)Using the modified GM(1,1)model to simulate the anchor PS curve,the residual of the anchor rod is obtained at each point,the data of each anchor is used as the input of the neural network,and the final prediction result is The residual of the actual ultimate bearing capacity is used as the output of the neural network;the normalized method of the input matrix is improved,the input matrix of the neural network is processed by the row-normalized method,and the ultimate bearing capacity predicted by the grey neural network model is inversed.Normalize to get the final forecast.(4)Based on the structural characteristics of the anchor rod,an experimental model of the ultrasonic guided wave excitation bolt anchorage is established,and the time difference of the reflected wave packet propagating in the anchor wave is calculated according to the propagation law of the ultrasonic guided wave in the anchor rod,and then obtained.The bolt anchorage section length.The relationship of displacement response of the anchorage end surface of the anchor under ultrasonic transient wave excitation was deduced,and the bond stiffness was brought into the formula of the relationship between the length of the anchorage and the shear stress,and the anchor rod was found to be in the state of elastic deformation and limit drawing.The bearing capacity was finally verified by a pull test.