International Journal of Geomechanics

When strong rock masses, with discontinuity patterns parallel and perpendicular to the ground surface, are subjected to normal loads, linear or concave stress-deformation curves are produced. In contrast, chalk rock masses with the same discontinuity pattern, produce convex curves. This paper investigates the underlying mechanisms, which may be responsible for such differences. Experimental results are presented for profiled chalk specimens in which the contact area at the discontinuity boundary is approximately 15% of the specimen cross-sectional area. It was found that these low contact area specimens exhibited both concave and convex behavior. This behavior was attributed to discontinuity closure and yielding of the intact material, respectively. The overall trend in behavior was found to be a function of the contact area at the discontinuity boundary, the initial discontinuity aperture, and the yield stress.

Response of a pile subjected to torsional loading is significantly affected by soil profile. To investigate this characteristic, pile-soil interaction may be modeled by a series of independent springs distributed along the pile shaft (load transfer model). With this modeling, in this paper, analytical solutions are developed for the pile embedded in a two-layered soil profile. The soil stiffness of each layer is assumed to increase as a simple power law of depth, starting with any value at ground level. The solutions are expressed in terms of Bessel functions of noninteger order; and also presented in simplified expressions for some special cases. The solutions are verified using a spreadsheet, numerical program GASPILE, which is based on a nonlinear load transfer model. They are further collaborated using more rigorous numerical approaches available for a two-layered homogeneous soil. With the solutions, extensive parametric study was conducted and presented in nondimensional charts, in order to show consistency among these solutions, significant influence of stiffness profile on torsional response, and to identify conditions of utilizing these solutions. A case study is provided to illustrate the use of the current solutions, in particular for deducing soil modulus from torsional test results.

Nowadays most hollow cylinder apparatus (HCA) in the world can only apply two dynamic loads, namely the axial load $\left(W\right)$ and torque $\left(M_T\right)$ , during cyclic principal stress rotation. The limitation of this loading mode is presented, and based on such limitation other possible and applicable loading modes are put forward: (1) On the premise of setting $W$ a special constant related to the inner $\left(p_i\right)$ and outer $\left(p_o\right)$ cell pressure, the stress path at which shear stress $\left(q\right)$ regularly changes with rotation angle $\left(\alpha \right)$ under the constant $b$ (Bishop parameter) and the constant mean principal stress $p$ can be achieved by adjusting $M_T$ ; (2) setting $b$ at a constant (not equal to 0.5), the linear relation between $p$ and $q$ can be realized under a regular changing $\alpha$ . In this case $W$ and $M_T$ should satisfy a certain equation and meanwhile the peak value of $q$ should be restricted within the $p-q$ critical state line; (3) without regarding the effect of intermediate principal stress, a stress path of $({\sigma }_1+{\sigma }_3)∕2$ monotonously changing with $\alpha$ can be achieved under constant $q$ . This implies that $W$ and $M_T$ should satisfy a combination of a number of simple trigonometric functions; and (4) the stress path of constant $b$ (with a random value between 0 and 0.5) and monotonous relation between $p$ and $R$ can be available under a regular changing $\alpha$ . In realizing this stress path, $W$ and $M_T$ should satisfy certain equations and the maximum $R$ should also be restricted by the critical state line. Results of this research can not only validate possible stress paths in cyclic principal stress rotation tests but can also provide a basis for further HCA improvement.

In the study of time dependent behavior of rock, the main difficulty is to predict delayed failure, which is of the utmost importance in assessing the safety of underground structures, such as deep underground facilities designed for high-level radioactive waste disposal. In this context, the viscoplastic behavior associated with the rock damage must be taken into account. As the longitudinal and transversal wave velocities are related to the physical and mechanical characteristics of materials, ultrasonic measurements can give valuable information about the development of damage. In this study, P-wave velocity measurements were used to monitor damage evolution during uniaxial strain in controlled compression tests and long-term creep tests. These measurements were performed using sensors in a piezoelectric copolymer of polyvinyl-difluoride, which were placed on both ends of cylindrical rock specimens. Throughout the experiments, the dilating behavior of an argillite could be correlated with a decrease of the P-wave velocity. Our results show that during a creep test, P-wave velocity measurements allow the three different phases of creep to be distinguished. During primary creep the P wave increases because of pore closure. The secondary creep phase, characterized by a constant strain rate, is identified by a linear decrease of the wave velocity; this trend accelerates during tertiary creep.

It is commonly assumed that within inundated sand the Skempton $B$ value and P-wave velocity decrease with decrease in saturation. In centrifuge tests a common saturation procedure is to inundate the specimen with carbon dioxide while under a vacuum and then slowly introduce the viscous pore fluid. The $B$ value and related saturation is difficult to measure in centrifuge models and P-wave velocity—saturation correlations have been used for this purpose. A laboratory emulation of centrifuge saturation procedures was made using a triaxial cell with top and bottom bender elements and a viscous methyl cellulose–water pore fluid. Contrary to expectations, the laboratory tests showed high P-wave velocities indicative of full saturation when $B$ values were low. Numerical modeling of the laboratory tests indicated that if air bubbles within the pore fluid are numerous and closely spaced then there is a good correlation between saturation, $B$ value, and P-wave velocity. However if the air bubbles are larger and only present in some of the pores then the P-wave velocity is not a good indicator of $B$ value and average saturation. The laboratory tests also showed that placing the specimen under backpressure for several days increased saturation and related $B$ values. It is suggested that this common laboratory procedure should be considered for saturating centrifuge test specimens.

The reliability of in situ compaction measurements made in producing gas/oil reservoirs with the aid of the radioactive marker technique (RMT) is a very important issue as an essential parameter controlling the geomechanical response of fluid extraction from the subsurface, i.e., the rock vertical compressibility $c_M$ , is obtained from RMT data. The validity of $c_M$ as provided by RMT has recently been questioned on the basis of the possible disturbance that a fully cased hole can induce on the actual reservoir compaction. By the use of a numerical approach and real casing information, the present paper addresses and discusses the reliability of RMT measurements performed in hole casings of the Northern Adriatic gas fields, Italy. The results show that the casing stiffness affects the RMT measurements only marginally, and especially so at a larger depth (i.e., $\mathrm{3,000}\mathrm{m}$ ), and does not significantly impact on the estimate of $c_M$ derived from RMT data.