PSI - Issue 10

D. Kotsanis et al. / Procedia Structural Integrity 10 (2018) 112–119

113

D. Kotsanis et al. / Structural Integrity Procedia 00 (2018) 000 – 000

number of failure criteria have been developed and important mechanical parameters have been calculated, (Paterson and Wong (2005)). According to the recommendations of the International Society for Rock Mechanics, there are three ways to conduct a conventional triaxial test, (Kovari et al. (1983)):  the single stage triaxial test (type I),  the multiple failure state triaxial test (type II) and  the continuous failure stage test (type III). With the single stage test, several triaxial tests are required in order to determine the strength envelope of a rock sample with satisfactory accuracy, over the required range of confining pressures, where the actual number of speci - mens needed depends on the intrinsic variability of the rock and the scatter of the data. As an alternative, a methodo logy was proposed by Kovari & Tisa (1975), called the “M ultiple Failure State Triaxial T est (or simply the “multi -stage triaxial test”), in which, by using only one specimen, more than one stress points of the failure envelope are obtained. In the present study, a number of multi-stage triaxial tests were performed on representative specimens from Alpine (calcite marbles, calcite brecciated marbles) and Post Alpine formations of Neogene age (sandstones, breccia and limestones) from the East Attica Prefecture of Greece. The peak and the residual strength envelopes of these specimens were obtained and their strength parameters were determined through simple regression techniques. The emphasis of this work is to compare the applicability of two different failure criteria in order to model the behaviour of intact rocks under multi-stage triaxial testing for the aforementioned states of stress. This study, that is part of an ongoing doctorate thesis, has employed an extensive laboratory program for the determination of the basic physical and mechanical characteristics of the rocky formations of the East Attica Prefecture of Greece. In this investigation, a total number of 31 intact rock specimens were prepared from selected rock blocks from natural outcrops of the study region and were tested under multi-stage triaxial stress conditions. Three petrological types were used and distinguished as (a) calcite marbles and calcite brecciated marbles, which belong to the Alpine para- autochthonous and allochthonous geotectonic units of Attica, (b) compact and porous to cavernous sparitic limestones as well as micritic limestones of lacustrine origin related to post Alpine sedimentation, (c) clastic rocks and in particular sandstones and breccia, which are also related to the aforementioned Post Alpine sedimentation process of Neogene age. A servo-hydraulic driven loading frame with a capacity of 5000 kN was used for the application and measurement of the axial load applied to each rock specimen. Two linear variable differential transducers (LVDTs) with a travel distance of 10 mm were placed in opposite directions between the loading platens in order to measure the axial dis placement. A Hoek-Franklin triaxial cell after Hoek and Franklin (1968), connected to a servo-hydraulic control unit, was used in order to apply the confining pressure ( σ 2 = σ 3 = p ) to the specimens. The preparation of the specimens was in accordance with the recommendations of the ISRM (Kovari et al. (1983)) and the slenderness ratio was approxi mately 2:1, with diameter equal to 54.5 mm. For each test, a number of 4 to 7 loading stages were achieved, while the final confining stress never exceeded 35.5 MPa, in the case of the hard rocks. As schematically illustrated in Fig. 1, at each stage the axial load was increased under a constant axial displacement rate, while the confining stress was kept constant until the imminent failure point has been reached. The confining stress was then increased to a higher level. This step was followed by an axial load increase until the next imminent point of failure was reached. This procedure was repeated until the maximum desirable confining pressure was achieved. Then, the axial displacement was increased until it reached the peak strength of the specimen. At this point the peak strength envelope of the specimen for the multi-stage triaxial test could be obtained. Loading of the specimen was continued with reduced axial load until it reached the residual state for the maximum confining pressure. The residual strength was determined as the stress where a flattening tendency of the load-displacement curve was achieved. Afterwards, the confining stress was progressively reduced with a constant rate of 0.02 MPa/s until the specimen was unloaded. According to the ISRM suggested method (Kovari et al. (1983)), the residual strength envelope could be obtained from this final stage. 2. General characterization of tested rock types 3. Experimental procedure