PSI - Issue 3

S.K. Kourkoulis et al. / Procedia Structural Integrity 3 (2017) 326–333 S.K. Kourkoulis, D. Triantis, I. Stavrakas, E.D. Pasiou and I. Dakanali / Structural Integrity Procedia 00 (2017) 000–000 5

330

typical load-time curve, which is equivalent to a load-displacement one (given that the tests are implemented under dis placement- control mode) is plotted in Fig.4a for a specimen with a=4 cm. The perfect linearity of the plot is striking. In Fig.4a data regarding the time evolution of NMOD, as recorded by the two clip-gauges, are, also, plotted. It is observed that for about 90% of the tests’ duration the NMOD-time relation is linear for both clips. After this instant one of the plots starts increasing rapidly while the second one starts decreasing. It is thus indicated that only one of the two notches continues opening after this critical time instant, the one from which crack propagation will eventually start (as it was verified by the photos of the UHSC). The second notch starts closing. This behavior was common for almost all experiments, independently of the length of the notches: The fracture starts from the tip of one of the two notches (either the left or the right one) and then it propagates along an arbitrary path until it reaches the other side of the specimen. In most cases the crack did not end at the tip of the other notch (see the photos of Fig.3). The corresponding conventional fracture stress (defined as force over the load bearing area) was calculated as: where w is the overall width of the specimens’ gauge length. The mean values of the fracture stress and also of the maximum NMOD obtained for each group of specimens are summarized in Table 1. Typical stress-NMOD curves, as determined by both the two clip-gauges and the DIC technique are plotted in Fig.4b for a specimen with a=8 cm. Generally speaking, the results for the mechanical response of the specimens are in good agreement with previous ones published by Kourkoulis et al. (2006) and Vayas et al. (2009), taking of course into account the wide discrepancies reported in literature (Theocaris & Coroneos 1979, Zambas 2004) for the mechanical properties of Dionysos marble. Table 1. The fracture stress and the critical NMOD for the four classes of specimens. a=2cm a=4cm a=6cm a=8cm Fracture stress [MPa] 4.35 4.41 6.10 5.91 NMOD [μm] 19.8 28.1 29 31   w 2a d P fr   σ  (1)

(a)

(b)

8

32

6.6

Load Left clip Right clip

6

24

4.4

NMOD [μm]

4

16

Left clip Right clip Left_DIC Right_DIC

2.2

Load [kN]

Stress [MPa]

2

8

0

0

0.0

0

120

240

360

480

-10

0

10

20

30

Time [s]

NMOD [μm]

Fig. 4. (a) Typical load and NMODs vs. time curves for a specimen with a=4 cm; (b) Typical stress-NMODs curves for a specimen with a=8 cm, as obtained from the clip-gauges and the DIC technique. The agreement is quite satisfactory considering the brittleness of Dionysos marble. 3.2. Acoustic emissions and pressure stimulated currents It was mentioned earlier that, according to the data for the NMOD, it seems that typical specimens appear entering a critical state at about 90% of the maximum load attained. Motivated by the need to further verify this conclusion, the time variation of the electric signal recorded near the tip of the notch from which fracture started, was plotted in Fig.5, in conjunction with the cumulative hits from the respective acoustic sensor. Despite some peaks of the PSC values recorded, it can be said that the electric signal increases smoothly following an almost constant rate until t~430 s