外文翻译-锚固岩石节理三维特性的实验和数值研究

翻译部分英文原文3D Behaviour of bolted rock joints:experimental and numerical studyG. GrasselliDepartment of Earth Science and Engineering,Imperial College of Science,Technology and Medicine, London SW7 2BP,UKAccepted 29 June 2004 Available online 16 September2004Abstract: Rock bolting is the most effective and also the most economical means of supporting excavations in rock. Various types of bolts are used today,and an understanding of the way in which these bolts work is essential for an optimal,safe,and economical use.Fully grouted,untensioned bolts have been commonly used in rock mechanics(i.e.,mines,rock fall stabilisation,underground works)for many years.In the 1980s a new type of bolt,called Swellex,was developed,becoming more and more widespread because of their easy and fast installation. However,regardless of the type,the mechanical behaviour of the bolted rock jointis not fully understood,and only the experience accumulated on rock bolting gives the know-how for the reinforcement calculation and execution.In this paper the different mechanical responses of full steel bars as opposed to the frictional Swellex are discussed. The study was done through experimental tests coupled with numerical simulations.The analysis of the results obtained both from nite element(FEM) modelling,and from large-scale(1:1) shear tests on rock joints,reinforced with fully grouted rods and Swellex bolts, clearly shows that the two bolt types deform in dissimilar ways,responding very differently to shear load.Published by Elsevier Ltd.Keywords: Rock bolt; Swellex; Shear tests; Experimental tests1. IntroductionExcavation size is a key parameter in the stability of underground openings in rock, and the larger the excavation, the less stable the roof tends to be. In very poor rock masses, large excavations have an unsupported stand-up time less than the minimum time required to support the roof after the excavation. Therefore, the principal objective in the design of a support system is to help the rock mass to support itself. During recent decades the use of rock bolts, which actually form a part of the rock mass, has become more and more widespread. Indeed, the use of rock bolts is a very exible method that can be combined with wire mesh, shotcrete and concrete lining to cope with almost any situation encountered during mining or tunnelling.According to Kovari, the history of rock bolting began with a patent specication (no. 302909) submitted by Stephan, Frohlich and Klupfel in 1913. However, it was only in the 1940s that the use of rock bolts became common in the American mining industry. Although anchors have been used for few projects in North America and Europe in the early 1950s, it was in Australia with the Snowy Mountains Scheme Project that, for the rst time, a series of studies have been undertaken for a better scientic understanding of the mechanics of rock bolting, establishing anchors as method of permanent rock support in civil engineering. In Europe, rock bolting began to be used extensively only when the rst tunnels were built according to the NewAustrian Tunnelling Method (NATM) in the 1960s. During the 1970s and 1980s a number of researchers carried out in-situ tests for monitoring the behaviour of installed rock bolts, and the use of rock bolting became a standard method for rock mass support. In those years a wide variety of dowels and bolt types were developed to meet different needs which arise in mining and civil engineering.In the 1990s Lunardi showed that, instead of relying on supporting the ground following excavation, the use of rock bolts for systematic pre-reinforcement of the front of the tunnel increases rock strength prior to excavation, increasing the stability and safety of the underground works. Today, especially in tunnelling, the reinforcement of jointed rock masses with anchors is widely used because of its proved efcacy and its low cost. However, the exact mechanism of resistance of the bolted system is difcult to analyse, and this is due to three aspects: the presence of a discontinuity with its roughness, the presence of several different materials (bolt, rock, grout or resin) with different mecanical behaviours, and the three-dimensional aspect of the system that is hard to simulate (i.e., bolt typology, bolt orientation, and load conditions). It is certain that the bolts increase the rock mass resistance, but the magnitude of this increase is hard to calculate. As a consequence, there is no unanimously approved method for the design of the reinforcement system, and most of the techniques are left to the practical experience of the engineer.2. Past studiesMany experimental tests have been performed in order to study the mechanical behaviour of bolted rock joints. A number of researchers carried out experimental programs on different types of rocks, reinforced by various elements. A wide variety of bolt and dowel types have been developed, indeed, to meet different needs, which arise in mining and civil engineering. It is possible to classify the bolts as active reinforcements or as passive ones. The essential difference between these systems is that tensioned rock-bolts apply a positive force to the rock, whereas dowels depend upon movement in the rock to activate the reinforcing action.Active rock bolts generally consist of plain steel rods with a mechanical anchor at one end and a faceplate with nut at the other. They are always tensioned after installation. For short term applications the bolts are generally left ungrouted, whereas for more permanent applications, or in rock in which corrosive groundwater is present, the space between the bolt and the rock is lled with cement or resin grout.Passive rock bolts generally consist of steel bars, which are grouted into the rock (Fig. 1). Tensioning is not possible and the load in the dowels is generated by movements in the rock mass. In order to be effective, dowels need to be installed before signicant movement in the rock mass has taken place. Another common type of passive bolt is the frictional one, and among them, the most popular are the Split Set (for mining applications) and the Swellex (mostly used in civil engineering applications).Fig.1. Different types of passive bolts.No matter which type, rock bolts in situ tend to restrain the deformation of rock with an increase in their axial loads. In other words, it is the rock deformation that applies a load to rock bolts in situ. The opening of a rock joint, for example, applies a tensile load to both sides of the section of the bolt intersecting the joint. During joint opening, decoupling of the bolt interface is activated rstly at the joint, and then propagates along the interface with an increase in the opening displacement. Hence, pullout tests are usually used to examine the anchoring capacity of rock bolts.A great number of pullout tests have been conducted so far in various types of rocks. According to Li, during a pullout test the shear strength of the interface comprises three components: adhesion, mechanical interlock and friction that are lost sequentially as the compatibility of deformation is lost across the interface. The result is a progressive decoupling front that attenuates at an increasing distance from the point of the applied load. Based on experimental results, Li suggested that the decoupling front rst mobilises the adhesive component of strength, then the mechanical interlock component, and nally the frictional component. The shear strength of the interface decreases during this process, and when fully grouted bolts are subjected to a pull load, failure occurs at the weakest interface (bolt/grout, grout/rock).Freeman, monitoring the loading process of the bolts and the distribution of stresses along fully grouted rock bolts in the Kielder experimental tunnel, proposed the concepts of neutral point, pick-up length and anchor length. At the neutral point, the shear stress at the interface between the bolt and the grout medium is zero, while the tensile axial load of the bolthas a peak value. The pick-up length refers to the section of the bolt from the near end of the bolt (on the tunnel wall) to the neutral point. The shear stresses on this section of the bolt pick up the load from the rock and drag the bolt towards the tunnel. The anchor length refers to the section of the bolt from the neutral point to the far end of the bolt (its seating deep in the rock). The shear stresses on this length of the bolt anchor the bolt to the rock. These concepts clearly outline the behaviour of fully grouted rock bolts in a deformed rock formation when it is subjected to only axial load.Stillborg improved the concept of pull-out test, and he carried out a number of tests on rock-bolts and dowels installed across a simulated joint, using two blocks of high strength reinforced concrete. This type of test gives a more accurate representation of conditions encountered underground than does a standard pull-out test. Testing the Swellex, Stillborg showed that, opening of the joint concentrates loading onto the portion of the dowel crossing the joint, causing a reduction in diameter and a progressive de-bonding of the dowel away from the joint. The ductile characteristics of the steel allows the de-bonded section to deform under constant load until, eventually, failure. Following this research, Li developed two analytical models for rock bolts in situ, one for uniform rock deformation and another for discrete joint opening (Fig. 2).Despite the good results, the mentioned approaches did not consider the shear force mobilised in the bolt itself, nor the deformation of the bolt near the joint. Few analytical expressions have been developed to predict the behaviour of a bolted rock joint under shear stress, and not only axial stress (pull out).In the 1980s, Dight proposed an expression to predict the maximum force mobilised in the bolt, as well as the associated displacement of the joint: the failure of the bolt is determined as acombination of axial and shear forces, and the displacement is computed taking into account the yield of the grout. Based on Dights work, Holmberg and Stille proposed a method that gives a good prediction of the maximum bolt contribution when the bolt is inclined to the joint. An empirical expression to compute the maximum bolt contribution and the associated joint displacement was introduced by Spang. Numerical approaches to compute the ultimate resistance of the bolt under axial and shear loads were proposed by Pellet and Ferrero.However, despite these studies, the deformational mechanism of the bolt is not completely understood. The aim of this paper is to contribute to a better understanding of the inuence of several parameters involved in the phenomena (e.g., number of bolts, bolt section, angle between bolt and normal to the joint), compare the behaviour of different types of bolts observed during large scale laboratory tests, and discuss three-dimensional nite element (FEM) simulations carried out with the commercial code Z-Soil 3D.3. Experimental set-upMost of shear tests on bolted joints available in the literature have the shortcomings of having been executed on samples whose dimensions were restricted to laboratory size, and therefore were limited in the diameter of the tested bolts. In order to overcome this problem, a new laboratory set-up was designed with the aim of reproducing the conditions that engineers have to face in underground works. The experimental set-up consists of a large-scale shear test with a double system of symmetrical joints, formed by a system of three large blocks (100 60 60 cm each), symmetrically reinforced (Fig. 3).The blocks were made of concrete (Fig. 4), and the two joints were macroscopically smooth in order to avoid further uncertainties associated with joint roughness that are difcult to precisely consider.The aim of this research was to investigate the mechanics of passive bolts that are commonly used in road tunnels. Therefore, neither resin grouted rebar that are more expensive than cement grouted, nor Split-Sets, that are mostly used for mining applications, were used. Two types of bolts were tested and discussed in this paper: fully cement grouted 16 and 20 mm diameter bolts, and a frictional hollow bolt (Swellex). In particular, the Swellex dowel, developed and marketed by Atlas Copco, consists of a 41 mm diameter tube, which can be inserted into a 3239 mm diameter hole (Fig. 2).The blocks were drilled at different inclinations ( =0,15 ，30，45) with holes of diameterdouble (32 or 40 mm) that of the rebar section to be installed (16 or 20 mm), and of 36 mm for Swellex, following the suggestion given by Atlas-Copco. The mechanical properties of the concrete, grout and each type of bolt were determined through standard laboratory tests. To measure the basic-friction angle on the joint, direct shear tests on an unbolted system were conducted for each set of blocks. The mean values are reported in Tables 1 and 2.In tunnels, the constraints of the surrounding rock mass may lead to increased the normal stress if dilatant shearing occurs. Hence, the most suitable condition to simulate that behaviour was to run shear tests at constant normal stiffness (CNS), that was assured by applying the normal load to the joint-system with two pre-stressed steel frames, which work as extremely stiff springs (K =210 GPa). The shear load was applied to the system pushing the central block progressively downwards. During each test the following parameters were recorded: the shear force, the force normal to the joint, the vertical displacement of the central block, and the deformation of the steel bolts.4. Discussion of tests resultsIn order to compare tests done with different types of reinforcements, we use thedimensionless bolt contribution to shear strength, T * , dened as the difference between theapplied shear force, Tv, and the frictional strength provided by the unbolted smooth joints,N tanfi ，normalised to the ultimate tensile load of the bolt,Fmax :T * = bolts contribution = Tv -2N tanfi2nFmax2nFmaxwhere N is the force that acts normally to the joint, i is the frictional angle of the smooth joint, and n is the number of bolt sections on each joint. The experimental results are reported inTable 3.The observation of the experimental curves shows similar shapes for both the full steel bolts and the frictional hollow ones. Fig. 5 compares the typical behaviour of two different types of reinforcement: full steel bolt and hollow ones. Regardless of the type, each experimental plot can be seen to have three different parts, each one corresponding to different mechanical reaction of the reinforcement. However, depending on the bolt type, each step has a different weight on the nal behaviour.The rst part corresponds to linear behaviour, with small displacements and great increase of load. The full steel bolt mobilises 75% of its resistant contribution associated with a 1.8 mmdisplacement. In this stage the Swellex mobilises only 20% of its Fmax for a displacement of 0.5 mm. The steel deforms elastically.In the second part of the curve, a non-linear behaviour is observed that corresponds to the yielding of the materials. However, because of the different shape of the bolts, two different phenomena, depending on the kind of reinforcement, occur. The resistance contribution T*reaches its maximum value (about 90% of Fmax ).The third part of the curve corresponds to a nearly unconstrained plastic deformation of the bolt, until its failure. However, the failure mechanisms are different, depending on the bolt type and on the angle joint/bolt.Experimental results show that the resistance mobilised by the bolts is proportional to the total steel section. Therefore, two bolts provide a resistance double that of to one bolt (Fig. 6). Indeed, the experimental tests point out that the variation of the bolt inclination affects both the maximum load mobilised by the reinforcements and the rigidity of the jointed system (Fig. 7) .The distance between the hinges was measured directly on the bolts extracted from the blocks after the test (Fig. 8), and the angle of the bolt at failure was calculated (Table 4). A linearly decreasing tendency with increasing the initial inclination angle between bolt and joint can be noticed ( Fig. 9). The extrapolation of scattered hinges distance data with a quadratic equation suggests that for an initial position angle =8085 there will not be any hinge and the bolt should work only in traction (Fig. 10). This conclusion seems reasonable, because the reinforcement is almost parallel to the joint.Fig.6.Inuence ofthe number of bolts on the mechanical behavior of reinforced joints.试验（deg） (deg) 平均值（deg）误差（deg）AB（cm）AB 平均值（cm）误差AB（cm）1A0505.01A0445.62A04744.53.44.84.70.52A0373.22B15346.02B154339.83.85.45.50.43B15435.03B15395.41C30254.01C30165.01C30305.21C30234.42C30253.82C30383.02C302926.81.55.64.80.22C30333.23C30236.63C30335.23C30256.03C30215.61D45155.61D451918.31.13.44.70.31D45184.02D45215.6Table 4Distance between hinges and angle at failure measured on bo