PSI - Issue 28

Dimos Triantis et al. / Procedia Structural Integrity 28 (2020) 502–510 4 D. Triantis, I. Stavrakas, A. Kyriazopoulos, E. D. Pasiou, S. K. Kourkoulis / Structural Integrity Procedia 00 (2019) 000–000

505

mixture was cast into moulds of prismatic shape (made of metallic material) with oiled internal surfaces. To achieve improved compaction and remove trapped air, a desktop vibrator was used. After 24 hours the specimens were removed from the moulds and were stored in a chamber at a constant temperature 22 °C and humidity equal to about 80%. The specimens were of square cross section (50 mm x 50 mm) and their height was equal to 70 mm. 3.2. Experimental procedure According to the protocol described here the specimens were submitted to monotonic uniaxial compression until fracture. Teflon sheets, interposed between the specimens’ bases and the loading platens, ensured both electric insulation and, also, minimization of friction. Load control conditions were adopted for all tests, at a rate of 70 kPa/s corresponding to a quasi-static loading scheme. A pair of electrodes, properly attached on two opposite lateral surfaces of the specimens at their mid-section (in such a way that the fictitious line connecting their centroids to be perpendicular to the loading line) were used to detect the electric currents with the aid of a Keithley 6514 electrometer, as it is shown schematically in Fig.1. In addition, an axial strain gauge, glued, also, at the specimens’ mid-section was used to record the axial strains developed, with the aid of a Kyowa bridge. The quantities recorded during the tests were the applied load, the axial strain and the PSC. Four classes of specimens were tested, at the 3 rd , 14 th , 28 th and 90 th day after casting. For each class three to five specimens were tested depending on the scattering of the data.

Fig. 1. The experimental set-up.

3.3 Raw experimental data The mechanical response of the four classes of specimens tested is shown in Fig.2, in which characteristic axial stress - axial strain curves are plotted. In accordance with what was expected the compressive strength increases with increasing age until that of twenty-eight days. The same is true for the stiffness of the material as it is expressed by the elastic modulus. Moreover, it is observed that the linear portion of the stress-strain curve for the specimens of age equal to three days is very short, compared to that of the remaining classes. Equally important is the observation that with increasing age the material becomes more brittle, as it is concluded by the gradually reduced “ductility” of the specimens, i.e. their strain at fracture. The raw data, concerning the evolution of the PSC developed during the loading procedure, are plotted in Fig.3, again for all four specimen classes tested. In this figure the PSCs recorded is plotted versus time which is normalized over the duration of each test (or more specifically versus the time at which macroscopic fracture occurred), in order for the plots to be comparable to each other. Ignoring the background electric noise, it is clear that for early ages (3 and 14 days) the PSC signal is quite stronger in comparison with that of mature specimens (28 and 90 days). The strong