Issue 28

A. Brotzu et alii, Frattura ed Integrità Strutturale, 28 (2014) 19-31; DOI: 10.3221/IGF-ESIS.28.03 25 Figure 9 : Instrument 13 inserted into the simulator. Figure 10 : Instrument 13 after the extraction test. The microcavities observed on the instrument surfaces have evolved producing microcracks since the early test cycles (Fig. 11). Often, these microcracks joint together producing cracks usually in the axial direction. However failure occurs only when this kind of defects develops on the file or in its proximity (Fig. 12). In this point the stress concentration is particularly high. Figure 11 : SEMmicrographs of instrument number 16. Figure 12 : SEMmicrographs of instrument number 15. The fracture surfaces are characterized by a wide central area with an extensive network of dimples, indicating a fracture due to overload (Fig 13). The fracture surfaces in proximity of the edges are usually highly worn. Suchmorphology is due to the fact that the instrument, after the failure, continues to rotate within the Plexiglas canal, until extraction. Thismakes it difficult to identify the failuremode in these areas. Taking into account that this problem does not allow to observe any signs of fatigue failure, some tests have been performed carefully removing the instrument from the simulacrum as soon as failure takes place, extracting from the plexiglas channel, the cutter broken tip and to observe it with SEM. This specimen part, when broken, does not rotate over and if the instrument is rapidly removed from the canal, it does not even suffers damages resulting from the rubbing of the two fracture surfaces. Themorphology of the fracture surface near the edges was studied on those specimens. Particularly, it has been observed on two well preserved specimens that near the cutting edges, dimples are not present, but typical failure striations due to fatigue can be clearly identified (Fig 14). It was alsopossible tomeasure the rate of crack propagation in these areas, which is very high, about 0.5 µm/cycle. The study of the load vs. displacement curves allows pointing out the effect of the lubricant gel. During the extraction test carried out without the use of gel the load rises linearly up to a value of about 27N. At this point, the tool is free and a sudden load drop can be observed. Then the load rises again, but slower than before, andwith a typical serrated curve up to the maximum load (approximately 32 N). This last load rise is due to the friction between the canal surface and the instrument. Instead, during extraction test carried out with gel, the load drop, which occurs at the detachment between instrument and canal, is never observed. Both the detachment and the maximum loads are lower than those measured