PSI - Issue 3

D. Triantis et al. / Procedia Structural Integrity 3 (2017) 346–353 4 D. Triantis, E.D. Pasiou, I. Stavrakas, I. Dakanali and S.K. Kourkoulis / Structural Integrity Procedia 00 (2017) 000–000 As far as it concerns the total electric charge released, which is calculated in terms of the test’s duration t f , as: f t T 0 Q PSC(t)dt   (1) it was found to be totally insensitive to the loading rate: It attains almost constant value, ranging between 281 and 295 pC, for all groups of specimens (Table 1). This conclusion supports further similar ones, previously drawn both theoretically and experimentally by Triantis et al. (2006) for marble specimens under uniaxial compression. The normalized PSC values recorded for characteristic specimens are plotted in Fig.3 versus the normalized load. Both quantities were normalized against their respective maximum value. From this figure, and also from Fig.2a, it is clearly seen that the way PSC varies depends strongly on the loading rate applied. More specific, it is observed that when the load is applied slowly (loading rates 28 N/s and 55 N/s) the variation of the electric signal is smooth (purple and blue curves in Figs.2a and 3a). On the contrary, for higher loading rates (88 N/s, 105 N/s and 135 N/s) the variation of the PSC is clearly more “wavy” (green, red and black curves in Figs.2a, 3b, 3c), while in addition peaks are observed, the number of which increases with increasing loading rate. Such peaks are probably related to early damage processes which in turn can be attributed to the more “violent” mechanical loading induced to the specimens. It is also noted that for low loading rates, PSC exceeds for the first time 50% of its maximum value when the respective load L 0.5 exceeds 75% of the fracture load, L f (Fig.3a). On the contrary, for higher loading rates, PSC reaches for the first time 50% of its maximum value when the respective load L 0.5 , is only equal to about 25% to 35% of L f (Figs.3b,c). The L 0.5 /L f ratios are recapitulated in Table 1 and Fig.3d for all five groups of specimens. 349

(a)

(b)

(d)

(c)

0.0 0.2 0.4 0.6 0.8 1.0

L 0.5 /L f

0

20 40 60 80 100 120 140

Loading rate (N/s)

Fig. 3. (a-c) The variation of the normalized PSC vs. the normalized load; (d) The load level at which PSC=0.5PSC max vs. loading rate.

3.3. Acoustic and electric emissions The micro-cracking activity is generally related to the acoustic hit rate, which was here calculated over the unit time, i.e. the number of acoustic hits recorded per second. The acoustic hit rates for characteristic specimens are presented versus time in Fig. 4a. It was found that considerable increase of the acoustic activity is observed at critical time instants, t cr , equal to t ct ≈80s, 45s, 30s, 25s, 20s, respectively, for the loading rates applied, in increasing order. Considering the overall duration, t f , of the respective tests, which were equal to t f ≈105s, 55s, 35s, 30s, 25s, and also the fact that the tests were carried out under load-control mode, it could be said that the acoustic activity is considerably amplified at load levels equal to about 80% of the maximum load attained (Fig. 4b). This character istic, if further verified by additional tests, could be considered as a criterion designating entrance to critical state. A second parameter, taking into account the amplitude of the acoustic emissions, which is widely used for the assessment of damage, is the b-value, a quantity adopted from seismology (Colombo et al. 2003; Rao & Lakschmi 2005). About fifteen years ago, the improved b (Ib) value was introduced by Shiotani et al. (2001) defined as follows: