ABSTRACT: The applicability of a new test method involving semi-circular (truncated) Brazilian disc specimens is examined in this study to effectively eliminate the unwanted high shear stresses at the loading points, as well as the multiple cracking that commonly occurs in the testing of standard Brazilian dicks comprising hard and brittle materials. A series of tests with a wide range of hard and brittle materials, including various types of rocks and ceramics, was carried out. In all tests performed, a single straight plane of fracture was observed. Through finite element numerical modelling and high-speed photography of the fracture process, the rupture mechanism was further confirmed to be the result of an indirect tensile stress induced inside the specimens. Based on the results, it is concluded that truncated Brazilian specimens can conveniently be adopted to determine the tensile strength of a target material, rather than the standard disc profiles in which catastrophic local cracking makes the procedure unreliable.
The Brazilian (splitting) test has been studied extensively in the literature since it was first introduced in the early 1940s. It is now a popular and preferred indirect method to measure the tensile strength of rock, pavement, concrete, and similar building materials. The test simply involves splitting a thin solid disc loaded by a compressive line load to generate a compression-induced tensile stress inside the disc until it fails across its loaded diameter into almost two equal hemi-cylinders (ISRM, 1978; ASTM, 2016). However, despite the test being widely accepted in geomechanics, test results with rock-like geo and construction materials having a large brittleness index (i.e. a high ratio of compressive to tensile strength, as in hard rocks, ceramic, alloys, and diamond composites) have been extensively criticised and proved not to be generally reliable (Fairhurst,1964; Hudson et al., 1972; Cranmer and Richerson, 1988; Yu et al., 2009; Swab et al., 2011; Serati, 2014). The main reason for the lack of reliability in testing such materials is the concentration of unwanted cracks developed in the vicinity of the contacts, which adversely interfere with the tensile breakage of disc specimens through the formation of triple-cleft fractures, inverse conical shear plugs, and multiple cracking (Ovri and Davies, 1987; Fahad, 1996; Serati et al., 2015). Therefore, the flexural tensile strength measured at the centre of a Brazilian disc often tends to overestimate the actual tensile strength in such hard and brittle engineering or natural materials. Nevertheless, this failure to achieve the standard breakage in the Brazilian test has been ignored in most cases since the reasons responsible for such catastrophic crushing are not yet fully understood.
ABSTRACT: The Brazilian test is one of the most established techniques used for indirect measurement of rock tensile strength. When the test is applied to soft to medium strength rocks, the rupture mechanism is always through the formation of a single diametrical crack initiated at the centre of the disc, where the induced tensile stress is the maximum. However, when applied to disc specimens of harder, stiffer and brittle rocks, secondary shear fractures often interfere adversely with the diametrical (characteristic) tensile crack in the Brazilian test. To prevent such undesirable shear cracking, a series of Brazilian tests on specimens of different sizes was carried out with four different brittle rocks, including granite, basalt and two types of monzonite. From the results, it is concluded that the problem associated with the undesirable shear failure developed in the vicinity of the contact points can be effectively eliminated through a change in the size of the disc in the Brazilian test.
The split cylinder test, also known as the Brazilian test, is the preferred laboratory technique for measuring the tensile strength of a wide range of rock-like materials. The test simply involves compressing a thin solid disc through the application of two diametrally opposite point (line) loads until the disc fails along its loaded diameter into two roughly equal halves (ISRM, 1978; ASTM, 2016). However, despite its ease of use and continued popularity, the validity of the test for materials having a high ratio of compressive to tensile strength, e.g. as in hard rocks, has been questioned in the literature by many researchers (Fairhurst,1964; Hudson et al., 1972; Cranmer and Richerson, 1988; Swab et al., 2011). The main concern is the type of failure reported for such materials, due to the typical formation of multiple cracks and shear ruptures in the vicinity of the contact points, rather than a single crack at the disc’s centre (Ovri and Davies, 1987; Serati et al., 2015). Obtaining a measure of the tensile strength where there is not a single central splitting crack is generally believed to be erroneous, unreliable and misleading. As partial solutions to this dillema, the Brazilian test has been continuously modified over the years, and new testing techniques have been proposed. These include development of the Ring test (Ripperger and Davids, 1947) and the flattened Brazilian disc test (Wang et al., 2003), as well as introducing loading arc platens for strip (distributed) loading of the Brazilian disc (Erarslan and Williams, 2012). All these tests have been shown to be able to successfully eliminate the shear fracturing and crack branching in the Brazilian test, but each also has its own limitations. For instance, the available closed-form solutions for the Ring test are often limited to simplified stress assumptions at the boundaries that may not properly represent the true contact conditions (Serati and Williams, 2015). The flattened Brazilian disc requires additional machining to prepare the flat ends, and the results are highly affected by the flatness and parallelness of the flat ends prepared.
ABSTRACT: The paper presents a numerical modelling approach by discontinuum method to simulate the progressive caving of top coal and roof above a Longwall Top Coal Caving (LTCC) face in Bowen basin, Queensland, Australia. A new caving modelling approach has been developed by the application of strainsoftening material model within UDEC in order to represent the failure and caving induced by the mining process. Results of numerical simulation have improved the understanding on LTCC mechanisms in terms of stress redistribution, rupture mode and failure mechanism in coal and roof strata. These results are in general agreement with site observation and previous longwall studies.
Longwall Top Coal Caving (LTCC) is an improved longwall mining technique that has shown many advantages compared with other methods in terms of coal recovery rate, face equipment design, development cost and spontaneous combustion control for extracting coal seam with thickness in excess of 4.5 m (Hebblewhite et al. 2002). Regarding the geotechnical issues, the applicability of LTCC is believed to highly depend on the cavability of top coal. The presence of top coal as a weak and highly jointed roof rock combined with the higher caving height (Fig. 1) makes the roof caving mechanism different from that caused by conventional longwall mining. Numerical modelling has been widely applied in LTCC studies due to its potential to represent the multiple caving mechanisms under various mining conditions. However, for continuum methods, the models have been limited to implicitly simulating the caving. The discontinuum methods, on the other hand, have mostly used the elastic model for intact rock and thus did not fully represent the failure behavior during caving. This paper aims to improve the understanding of mechanisms associated with caving by developing a discontinuous modelling approach using UDEC strain-softening model (Itasca Consulting Group 2014). The developed model is not only able to explicitly simulate the progressive caving but is also able to incorporate the plastic material response.
The conventional interpretation of nuclear magnetic resonance (NMR) measurements on fluid saturated reservoir rocks assumes that the T2 distribution and the pore size distribution are directly related. However, this relationship breaks down in many multi-scale porosity systems due to diffusion coupling between different pores. This case is more common in unconventional and carbonate rocks both because of smaller pore sizes and consequently shorter distances to be traveled by moving spins, in addition to significantly higher diffusion coefficients present in the case of gas reservoirs. A particular case not generally considered is that of inter-bed rather than intra- or inter-pore diffusion coupling. In such cases, traditional methods to estimate permeability and movable fluid fractions like the T2,cutoff method may give erroneous results.
This paper aims to illustrate through simulations and experiments the impact of inter-bed diffusion coupling on NMR responses of laminated porous media exhibiting different mean pore sizes. Two 3D structures have been modeled based on a Boolean particle process, providing a range of structural to diffusion length ratios to explore the relationship between pore geometry, surface magnetic properties, and NMR transverse relaxation time. Moreover, synthetic laminated systems of two different grain diameters with varied layer thickness were made and high-resolution micro-CT images obtained. Low field NMR experiments were carried out at 100% water saturation.
The magnetization decay affected by diffusion coupling was acquired numerically by running two different simulation sets, the first when coupling was allowed and the second when coupling was prohibited. This information was then used along with prior knowledge about responses from the individual layers in a careful treatment to identify the coupling strength (ξ) to improve the apparent pore size distributions. A coupling strength (ξ) is introduced as the difference in T2lm between the cases of coupling and non-coupling NMR responses of the individual thinbeds over the measurement time. The value of ξ represents the degree of diffusion coupling on a scale from 0 to 1 where 1 means total coupling between the thin layers. We calculate the vertical and horizontal permeability via the lattice Boltzmann method as references to study the improvement in NMR permeability estimation after completing the decoupling process.
The influence of surface relaxivity and diffusion coefficient on T2 relaxation responses have been tested to study the diffusion coupling strength. An escalation in pore coupling was observed with decreasing bed thickness as well as increasing the diffusion coefficient and decreasing the surface relaxivity. When pore coupling was strong, the T2 distribution clearly misrepresents the underlying bimodal distribution of the different morphologies. Consequently, the bimodal relaxation time becomes merged and the relative amplitude of the peaks fails to reflect the true morphologies of the models. The decoupling process by using the value of ξ was applied, rock-typing was successfully achieved, and horizontal and vertical permeabilities derived with good accuracy.
Copyright 2016, held jointly by the Society of Petrophysicists and Well Log Analysts (SPWLA) and the submitting authors. This paper was prepared for presentation at the SPWLA 57th Annual Logging Symposium held in Reykjavik, Iceland June 25-29, 2016. ABSTRACT Knowledge about rock wettability is critical for optimizing oil recovery from primary to tertiary development stage. NMR remains the best option as logging tool to infer several petrophysical properties including potentially wettability, with limitation, as it can work as a cross-link between special core analysis (SCAL) and well logging. However, so far neither NMR nor other techniques are capable to quantitatively evaluate wettability conditions of mixed-wet reservoirs. Furthermore, such conditions require corrections even for standard petrophysical interpretation of NMR responses. Conventional interpretation of NMR responses from partially saturated rocks relies on strong discrimination of wetting and non-wetting fluids due to their physical contact with the solid surface and hence very different effective surface relaxivities. Mixed wettability conditions complicate NMR interpretation due to diffusional exchange of a signal from an individual fluid phase which is partially in contact with solid phase of the rock and partially isolated by a nonwetting surface or fluid boundary. Several theoretical and experimental studies have been carried out to illustrate this behavior with the attempt to employ NMR as wettability indicator on a porous medium. This work presents an advanced workflow for the interpretation of NMR responses of mixed-wet microporous carbonate rocks by employing micro-CT images and forward simulations of NMR responses, providing spatial distribution and strength of wetting properties at mixed-wet conditions. Simulated NMR relaxation responses at different partial saturations and wetting states are validated by corresponding experimental data. The obtained promising simulation results may help to pave the way for more confident NMR interpretation of a broad range of wettability conditions and for better reservoir characterization and development.
Joints can significantly affect the mechanical behaviour of rock masses. The presence of a joint set is crucial to the initiation and propagation of caving. A numerical approach to cave assessment requires a realistic joint constitutive model, and therefore produces better prediction of the cavability of the orebody. An asperity degradation model for rock joints has been developed that considers the progressive abrasion of a true roughness area over joint sliding. The magnitude of dilatancy is predicted to be decreased exponentially with the increase in shear displacements. The degradation in dilation and post-peak strength along asperities is modelled on the basis of the wear process. Then, geometric conditions and rock strength are considered through the dimensional analysis. Experimental studies of direct shear tests have been conducted using triangular shaped asperities and the results are correlated with the model’s behaviour to demonstrate its performance. The proposed joint model can be readily implemented in numerical procedures such as discrete element method and used to simulate block or panel caving.
Block cave mining has drawn increasing attention in recent years because it permits the bulk extraction of low grade ore bodies in a cost effective manner. The cave propagation of a jointed rock mass is strongly governed by the frictional characteristics of the geological discontinuities and in particular, joints in rock. Therefore, explicit and accurate formulation towards the shear behaviour of joints avails prediction of rock mass caving.
Amongst many contributing factors, surface roughness plays a key role in the friction between joint walls. On one hand, roughness dilatancy serves as an important stabilizing effect. Two contacting bodies tend to separate during tangential movement due to the sliding of asperity surfaces of one body on the other. When the increase in contact surface volume is constrained, dilatancy augments normal stresses compressing joint walls, which in turn, can significantly increase a joint’s resistance. The asperity surfaces responsible for dilation, however, will degrade and affect the subsequent shear behaviour depending on the normal stress level and the mount of sliding.
Goldstein et al and Patton  are among those who first attempted to predict the shear strength of non-planar rock joints based on the dilation caused by asperities. Thenceforth, the dilative feature of rock fractures has been addressed in both empirical and theoretical approaches by numerous researchers such as Ladanyi and Archambault , Barton , Schneider , Leichnitz , Plesha , Jing et al, Wibowo  and Oh et al.
The joint surfaces are irregular in nature; the roughness degree has been interpreted using statistical and fractal approaches [11-19]. As joint asperities degrade in shear, the surface damage has been investigated by several researchers based on the aforementioned methods including Gentier et al, Homand et al, Grasselli and Egger , Belem et al and Jiang et al. It seems that complete descriptions on asperity degradation are absent from those models because only geometric changes are taken into account. However, other factors such as normal stress magnitude and material strength are closely associated with the surface damage [25-28].
This paper presents a tribological relationship for joint asperity degradation based on the theory of wear. By dimensional analysis, factors affecting asperity deterioration including the applied stress and joint strength are taken into account. Examples are considered comparing the proposed model to experimental data.
Since the introduction of the point load test, size effect has been observed in the point load strength index. While considerable research has been undertaken to investigate the size effect in the point load strength index, and the general cause of size effect, there has been limited research applying size effect theories to the size effect observed in the point load strength index. This paper investigates the applicability of size effect models including SEL, MFSL, Brook Model and FFSEL to both axial and diametral point load strength indexes at varying sample length to diameter ratios. In addition, the size effect at varying length to diameter ratios of a point load strength index which incorporates the contact area between the loading pointer and the sample has also been investigated.
Through an experimental investigation of both axial and diametral point load tests involving 374 samples of Gosford sandstone, this investigation found that the point load strength index varies increasingly with sample diameter for all investigated samples at different length to diameter ratios. The SEL and MFSL size effect models best fitted to the size effect trend of Gosford sandstone results obtained through axial point loading. However, when the contact area was included then the point load strength index found to increase with sample diameter for all investigated length to diameter ratios. The FFSEL size effect accurately modelled the increasing trend of Gosford sandstone results under both axial and diametral conditions across all length to diameter ratios.
The point load test is widely used within geotechnical and rock engineering in the classification of rock. Due to the low cost and portability of the test unit, it is commonly used within the mining industry to classify intact rock strength for use in rock mass classification systems such as the Rock Mass Rating (RMR) or Q systems . Knowledge of intact rock strength and the rock mass classification is a fundamental input for geotechnical design and also is a parameter for selecting mining method. Accurate classification of intact rock strength is critical for the accurate classification of rock mass which is necessary for the development of safe geotechnical and mine designs.
Since the introduction of the point load test, a size effect has been observed in the test results where the measured rock strength varies increasingly with sample size. This effect can have a significant impact on the classification of intact rock strength. Considerable research has been conducted to investigate the size effect in the point load test with previous research finding that point load results vary with sample diameter and sample length to diameter ratio [2-6]. This research has led to the introduction of sample size requirements for the point load test.
Coupon testing has been used to investigate stress corrosion cracking (SCC) of rockbolts in a laboratory based experimental program. This program has focused on the use of ASTM G39 three- and four-point bent beam coupon specimens, immersed in an acidified sodium chloride test solution containing hydrogen sulfide. The specimens were loaded to a range of stresses from zero, to just above the materials yield strength. It was found that only specimens above an applied stress of 580MPa failed by means of SCC, showing that rockbolts loaded to near their yield strength may be susceptible to SCC in-situ.
Stress corrosion cracking (SCC) is a failure mechanism that affects material through the growth of hairline cracks when exposed to a corrosive medium while under stress (Jones and Ricker, 1999). SCC has been found to affect rockbolts in the Australian underground coal mining industry and has the capacity to cause the catastrophic failure of ground support systems, potentially leading to falls of ground in underground excavations, injury or death to mine personnel, damage or loss of equipment, and loss of productivity through the loss of roadway access (Crosky et al, 2002).
Coupon testing utilises small scale samples that are intended to be indicative of the behaviour exhibited by a material in-situ (ASTM G39). The small scale, and usually high surface area to volume ratio of coupon specimens mean that they corrode at faster rate proportional to their mass, greatly reducing testing times. The small scale of the specimens allows them to be more easily prepared and handled, loaning them to mass manufacturing techniques.
A previous study conducted by Vandermaat et al (2012b) into the use of three- and four-point coupon testing in-situ in an underground coal mine failed to result in SCC of the specimens. This paper serves as a follow up laboratory investigation in order to understand why these specimens did not fail. Previous studies into the SCC in rockbolts carried out by Gamboa and Atrens (2003) determined that rockbolt steel has a critical stress threshold of approximately 90% of the material ultimate tensile strength (800-900MPa). This study will focus on this critical stress threshold as a possible explanation for the lack of failure in the in-situ specimens (Vandermaat, 2012b).
With the plethora of cable bolts on the market, geotechnical engineers can be baffled by the differences in performance claims from the manufacturers but perhaps more importantly they are sometimes at a loss in making an optimal selection matching behaviour with a particular site condition. To this end a new testing methodology has been developed that can cater for the varying types of cable bolt behaviours. The methodology is based on the laboratory short encapsulation pull test (LSEPT) as recommended in the British Standard and overcomes several deficiencies such as cable rotation. Of most significance however, is its inability to measure the post-peak residual performance of high load transfer capacity cables and consequently, there is an inference that these cables may be susceptible to sudden anchorage failure with relatively little yield. This paper details the results of anchorage performance tests of a range of cable bolts and the differences in performance with rock strength and drill hole diameter.
A research project was undertaken to develop a new laboratory-scale test apparatus to study the axial performance of fully grouted cable bolts that are used in the Australian underground coal mining industry. Work by Thomas (2012) and others in recent years found there were a number of deficiencies in the testing methodologies making it difficult to determine and compare the performance of the dozen or more different types of cable bolts on the market. This has meant that geotechnical practioners have been hamstrung in optimizing the design of ground support systems in underground environments that are increasingly more arduous.
A cable bolt is a flexible tendon consisting of a quantity of wound wires that are grouted in boreholes at defined distances between holes in order to provide ground reinforcement of excavations (Hutchinson & Diederichs, 1996). They were initially introduced into the underground hard rock mining industry in the 1960s (Thorne & Muller, 1964) and since the early 1970s have been brought to coal mining operations.
Originally, cables were only used as a temporary reinforcement element. One reason for this being many earlier of the earlier cables were made from discarded steel ropes that had very poor load transfer properties due to their smooth surface profile, lacking the equivalent to ribs found on rockbolts. Over subsequent years a number of modifications have been made to the basic plain strand cable, such as buttoned strand (Schmuck, 1979), double plain strand (Matthews, Tillmann & Worotnicki, 1983), epoxycoated strand (Dorsten, Hunt & Kent, 1984), fiberglass cable bolt (Mah, 1990), birdcaged strand (Hutchins et al, 1990), bulbed strand (Garford, 1990), and nutcaged strand cable bolts (Hyett et al, 1993). These changes to the cable surface geometry have been undertaken in an effort to improve the load transfer efficiency and anchorage capacity that has resulted in the more widespread use of cable bolts for permanent reinforcement.
Fully grouted cable bolts were first introduced into the mining industry in the 1960s and they have played an increasingly important role in ground support particularly in both underground coal and hard rock mining. While there has much research and development in cable bolt design, the failure of cable bolting system especially slippage at the cable/grout interface is still a significant issue being reliant on the load transfer between surrounding rock mass and cable bolts. In order to better understand this mechanism and evaluate the influence of relevant parameters on the load bearing capacity of cable bolts, various methods have been devised over time including single embedment length test, double embedment length test, and the “split-pipe pull/push” test. The Laboratory Short Encapsulation Pull Test (LSEPT) is perhaps the most recent development in testing methods to assess the load transfer behaviour at the cable/grout and grout/rock interfaces. However, recent work has identified a number of issues with this method as outlined in this paper. Firstly, the diameter effect of the rock sample which is used to confine the grouted cable bolt is studied since the traditional LSEPT in the British Standard is only limited to a relatively small diameter of 142 mm being approximately only three times the diameter of some of the larger cable bolts. As well, the influence of bearing plate on anchorage performance due to similar dimensions of the hole size is considered, proposing an appropriate parameter dimension for the bearing plate. A modified version of the test method is proposed and is presented in this paper. This paper outlines the results of a study on some of the design parameters that formed the basis of the new modified test design.