PSI - Issue 33

Irina A. Bannikova et al. / Procedia Structural Integrity 33 (2021) 357–364 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

362

6

air was accelerates the impactor along the barrel. After the impactor collides with the input bar, a compression wave is excited in the latter. The compression wave is transmitted through the bar to the sample, after which, due to the difference in sound impedances, the wave decays into the wave that passed through the sample into the output bar and into the reflected back into the input bar. With the help of a PMT, the fractoluminescence of sample No. 3 was recorded in the process of dynamic failure, see Fig. 7. The failure was accompanied by multiple pulses of fractoluminescence. This is typical for damage to failure transition where microcracks are merging into macro cracks leads to complete fragmentation of the sample. The form of the fracture process under dynamic loading was similar to that observed a sample No. 4 under quasi - static loading (Fig. 6).

B

Fig. 7. Fractoluminescence (signal – yellow) registration by a PMT in the process of dynamic loading of a cylindrical sample No. 3 (quartzite) tested on SHPB. B – a multiple impulses – failure of the sample. 4. Fragmentation analysis and signals analysis The fragments of the all specimens were identical in shape and size regardless of the loading conditions. Sample fragments were sieved across sieves system with characteristic size of the opening ( d* , mm), and weighed on the electronic balance HR-202i with an error of 0.0001 g. For example, see Fig. 8. The total number of fragments was counted on all sieves except the pallet with fragments with d* < 0.05 mm. On sieves, where there were too many fragments, only a part was weighed (but not less than 0.01 g) and their number was counted. Then, using the proportion, the approximate number of fragments of the entire sieve was calculated. On any sieve under consideration, there were fragments, the shapes and its sizes (elongated volumetric objects, flat objects and in the octahedron form) of which are characterized by similarity. In weight saving mode the fragments mass was 98 percent of the sample mass in case quasi-static loading and ~70 percent in case dynamic loading (here we need to work on the purity of the experiment yet). The Fig. 9, a are shows the cumulative fragments distributions for samples obtained under quasi-static (No. 1, No. 2, No. 4) and dynamic loading (No. 3). The curves have two slopes separated by inflection point. Each slope is described by a power law, which is typical for the failure of brittle materials by Astrom and et al. (2004), Bannikova and et al. (2016), Katsuragi and et al. (2004), Meibom and Balslev (1996), Davydova and Uvarov (2013), Davydova and et al. (2014), Davydova and et al. (2016). For large fragments (from sieves with a characteristic opening size of 1.6 < d* < 12 mm – I on Fig. 9, a ), the power in the law was the same for all samples, for smaller fragments (from sieves with a characteristic opening size 0.1 < d* ≤ 1.6 mm – II on Fig. 9, a ) the degrees were different. Part of the curve runs below (for samples No. 3 and No. 4 at similar loading pressure by dynamic and quasi-static loading, respectively). Most of the small fragments were formed as a result of the formation and failure of large fragments.