The Solution Mining Research Institute (SMRI) has recently embarked on inquiries into the effect cyclic loading (both mechanical and thermal) might have on salt. Some of this interest stems from the concept of using salt caverns as a storage medium for renewable energy projects (e.g., Compressed Air Energy Storage [CAES] with wind power) where daily pressure cycles in the cavern are conceivable as opposed to the seasonal cycles expected for typical natural gas storage projects. RESPEC and the Institut für Aufbereitung und Deponietechnik (Chair for Waste Disposal Technologies and Geomechanics) at Clausthal University of Technology (TUC) jointly executed a rock mechanics laboratory study using both facilities for performing triaxial cyclic loading creep tests on rock salt recovered from the Avery Island Mine in Louisiana, USA.
The cyclic triaxial creep tests were performed under various load paths including compression, extension, and compression/extension. In addition, the tests were performed under both dilative and non-dilative stress regimes. The cyclic compression creep data was compared to static creep tests performed under similar conditions and the results indicated that the cycling of the applied stress did not have a dramatic effect on the specimen’s creep rate as long as the applied stress was not in the dilational regime. Furthermore, the cyclic compression tests were compared to a numerically simulated creep test at the same stress and temperature conditions, and the comparison matched closely for the nondilational loading segment.
Günther, R.-M. (Institute for Geomechanics GmbH (IfG)) | Salzer, K. (Institute for Geomechanics GmbH (IfG)) | Popp, T. (Institute for Geomechanics GmbH (IfG)) | Lüdeling, C. (Institute for Geomechanics GmbH (IfG))
Actual problems in geotechnical design, e.g. of underground openings for radioactive waste repositories or high-pressure gas storages, require sophisticated constitutive models and consistent parameters for rock salt that facilitate reliable prognosis of stress-dependent deformation and associated damage from the initial excavation to long times. Fortunately in the long term the response of salt masses is governed by its steady state creep behavior. However, because in experiments the time necessary to reach true steady creep rates can last time periods of some few days to years, depending mainly on temperature, an innovative but simple creep testing approach is suggested. A series of multi-step tests with loading and un-loading cycles allow a more reliable estimate of stationary creep rates in a reasonable time schedule. In completion, the advanced strain-hardening approach of Günther/Salzer is used which describes all relevant deformation properties of rock salt, e.g. creep and damage induced rock failure, comprehensively within the scope of an unified creep approach. The capability of the combination of improved creep testing procedures and accompanied modelling is demonstrated by recalculating multi-step creep at different loading and temperature conditions. Thus reliable extrapolations relevant to in-situ creep rates (10-9 to 10-13 s-1) become possible.
A rock mass is neither a continuous medium nor a totally discrete medium, it is a kind of defect material which contains many cracks, joints and faults. The nonlinear deformation behavior of a rock mass is induced by the propagation and coalescence of cracks and joints under external loads. Therefore, it is of important for rock engineering to analyze the propagation and coalescence process of cracks existed in a rock mass under external loads. The study of crack initiation and propagation is important for the understanding of rock mass behaviour which, in turn, affects rock engineering applications, such as tunnels, foundations and slopes, as well as hydrocarbon and geothermal energy extraction. Cracking mechanisms can be studied experimentally in the laboratory or in the field, or numerically. In the present study, distinct element method (DEM) which is capable to model various discontinuities was employed to simulate crack initiation and propagation in a rock masse specimen containing a single open and closed flaw. Initially, a rock domain containing a closed flaw was considered to model crack propagation. The analysis was performed by sequential modelling. Firstly, a model containing a single closed or open flaw was used to verify the types of propagation shown by Park and Bobet (2009). In these analyses, both open and closed flaws were considered and analysed with different spatial distribution (i.e. flaw angle). After results verification, the effect of open flaw filling material on crack propagation was analysed numerically. This characteristic has not yet been studied in crack propagation studies. The results obtained from open and closed flaws were in good agreement with experimental ones. All cracks mentioned in experimental literatures such as wing (tensile) cracks, coplanar secondary (shear) cracks and oblique secondary cracks were modelled successfully using DEM which indicates method capability to model nonlinear behaviour of rock masses subjected to external loadings. The emphasize of the study is to investigate the effects of open flaws containing filling material on crack initiation and propagation. Weak material was modelled as filling material. The results showed that when flaw is filled with weak materials (as it encountered frequently in natural rock masses) the cracking pattern is quite different with open flaws. In these occasions, the crack propagation direction is different. This phenomenon could be described in terms of stress attenuation in weak filling material, as stress concentration in filling materials causes change in crack propagation directions as well as crack length.
True triaxial tests have been carried out in two quartz-rich, high porosity, sandstones, Coconino (n = 17.5%) and Bentheim (n = 24%) by maintaining constant but different σ2 and σ3 and raising σ1 until failure occurred (σ1, peak). For each constant σ3 level, σ2 was varied from test to test between σ2 = σ3 and σ2 = σ1. σ1,peak for a given σ3 increased with σ2, reached a maximum (up to 15% higher than when σ2 = σ3), and then declined so that when σ2 approached σ1, σ1, peak was about equal to its base value when σ2 = σ2. A separate series of tests was carried out using a novel loading path by maintaining constant Lode angle T (= 0°). This series of tests characterized the dependence of the octahedral shear stress at failure toct,f on the octahedral normal stress at failure σoct,f when Θ = 0°. The latter tests were used to obtain the necessary parameters employed in a three-invariant failure theory proposed by Rudnicki (2013). The theory was then applied to predicting the variation of σ1,peak with σ2 for a given σ3. The prediction reasonably replicated the typical ascending-then-descending σ1,peak vs. σ2 trend observed experimentally in both sandstones, confirming (with some limitations) the applicability of the Rudnicki’s (2013) theory.
Ability to induce complex, highly connected fracture networks, that can remain open during production, is the key to unlock permeability challenged shale gas plays. Within the time and pressure scale of hydraulic fracturing operations, it is difficult to create fracture complexity in ductile shales. However, when subjected to a high rate/pulse loading, rock might exhibit a brittle to ductile transition and a complex fracture network might be created. Along these lines, the concept of pulsed fracturing, that customizes the pressure-time behavior of a pulse source to create multiple fractures, is introduced. In this paper, an integrated 3D model that quantifies fracture initiation, growth, and coalescence due to initial and post-peak pulse loading is presented. The simulation involves a numerical algorithm that couples tensile/shear/compactive failure algorithms with dynamic fracture propagation and pore fluid pressure. Geomechanical modeling approach makes it possible to optimize pulsed fracturing for different shale plays. After constitutive model description and presentation of key aspects of the model, the model is employed to a reservoir dataset to evaluate pulsed fracturing as an alternative fracturing technique. The results show that, if designed accurately, pulsed fracturing could help trigger a ductile to brittle transition and can generate complex fracture networks.
The shear loading performance of fully grouted rock bolts is presented. A new experimental method, two types of rock bolts (D-Bolt and rebar), joint gap opening effect, and three types of host rock material were conducted. The bolts have similar loading capacity whilst the loading angle has obvious influence on the deformation capacity of the D-Bolt. The maximum total failure displacement of the D-Bolt decreases from 140 mm (0°) to 70 mm (40°, 60°, and 90°), whereas that of the rebar keeps a smooth increase from 29 to 53 mm. The deformation capacity of the D-Bolt is 270%–50%larger than that of the rebar with increasing loading angle. The bolt has a larger deformation capacity in comparatively soft rock than in hard rock. A slight increase trend of failure load with a decrease of rock strength can be observed with 90° loading angle. The bolt subjected to 60° or 90° loading is much stiffer. The total stiffness of the bolting system is contributed mainly by shear component as the stiffness in shear direction rises significantly from 20° to 90°. The displacement capacity of the bolts increased with the joint gap.
Laboratory experiments have been performed to simulate in situ and core compaction behavior of soft reservoir sandstones. Fine-grained synthetic sandstones have been manufactured under simulated in situ stress conditions with various cement contents. A systematic study shows that in situ compaction is close to elastic only under initial conditions, and that plasticity develops gradually during compaction. Two mechanisms control rock alteration as a result of stress release during coring: Cement bond breakage, which leads to softer and more stress sensitive core material at low stresses, and grain rearrangement into a denser packing, leading to permanently reduced porosity and hence increased stiffness compared to in situ behavior. The relative importance of these mechanisms depends on the degree of cementation, on the coring stress path, and on the stress level. Ultrasonic velocities have been measured and linked to time lapse seismic response of soft reservoirs and to the feasibility of predicting it with core material. Time dependent (creep) deformation has been observed, and appears related to the evolution of plastic strain.
Experiments have been performed with Mancos Shale under Brazilian tensile test conditions, addressing the effect of the angle between layer or bedding planes and the loading direction. A high-speed camera with digital image correlation software is used in combination with acoustic emission recording to monitor the fracture initiation and growth processes during loading. Although a clear anisotropy is observed in the variation of the P-wave velocity with the inclination angle, a significant effect on the Brazilian tensile strength is not observed. The mode of failure depends however on sample orientation. For all specimens, a main diametrical central fracture is induced first. It originates in the middle of the specimen and grows as a straight line or as a zig-zag line, depending on the orientation of the sample with respect to load direction. The zig-zag fracture is then a combination of a fracture along the weak direction and in other directions. Its evolution is an order of magnitude slower than that of a brittle straight diametrical central fracture.
This paper presents an experimental investigation on both monotonic strength and fatigue crack kinetics for Berea sandstone. It is found that for all specimens the Paris-Erdogan law is applicable for a wide range of the amplitudes of the stress intensity factor. The fatigue tests also indicate that there is a small-crack growth regime at the beginning stage, where the growth rate decreases as the crack propagates. The fracture kinetics for both the small-crack growth and the Paris regimes is subjected to a strong size effect. During the strength and fatigue tests, the damage process is examined by the digital image correlation method, and it is shown that the length of the fracture process zone under cyclic loading is about 60% larger than that under monotonic loading. In parallel with the experiments, a theoretical model is developed to explain the observed size effect on fatigue crack kinetics.
Existing laboratory methodologies for characterizing the pore volume compressibility of rocks are summarized. Special emphasis is placed on the two most common in the industry pore volume compressibility tests—uniaxial strain pore pressure depletion tests and uniaxial strain effective stress loading tests. We carefully overviewed a rigorous mathematical description of uniaxial deformation of porous rocks implemented in these tests, derivation of pore volume compressibility coefficients from stress-strain data, and assumptions made in these models. Many industry-relevant porous rocks demonstrate nonlinear stress-strain behavior. As a practical workflow for characterization of pore volume compressibility, we propose using piecewise linear approximations of loading diagrams with constant compressibility coefficients. The linearized model provides a suitable description of nonlinear rock response within certain limited intervals of loading trajectory. For the correct use of established pore volume compressibility properties in applications, it is important to validate that predictions are made for the equivalent loading paths, stress, and pressure levels as used in the linearization intervals. Several examples of mechanical rock behavior were considered, focusing on the end-member cases of high- and low-grain compressibility compared with bulk compressibility. We also discussed implications of some rock microstructures, including mudstones, on mechanical properties, to provide a reference for better interpretation of real rock behavior.