PSI - Issue 12

15

E 1 [ GPa ]

P.M. Giuliani et al. / Procedia Structural Integrity 12 (2018) 296–303 P.M. Giuliani, O. Giannini, R. Panciroli / Structural Integrity Procedia 00 (2018) 000–000 10

301

6

The research then moved to the second flax fibers (BioTex). These have been impregnated with the SuperSap Resin only. Results for the BioTex / SuperSap specimens show a high dispersion in terms of both initial modulus E 1 and final modulus E 2 , as clearly observable on the left chart of Figure 5. Such behavior, which is not in line with the results obtained using the other kind of flax fibers, is found to be strictly related to the location of the specimen on the original laminate. It is found that, moving transversally to the fibers, the elastic modulus increases while approaching the center of the laminate, to later decrease again to the original value. We comment that the laminate has been produced utilizing a fiber roll 250 mm with, transversally cut every 300mm, and packing the laminae one on top of the other. The external fibers of the roll are thus always located on the external portion of the laminate, and the center of the roll always corresponds to the center of the laminate. It has been then found that the density of the fibers varies moving along the width of the roll, with higher density at the mid-span. Whereby we can not give an exact estimation of the fiber density variation, it can be stated that the variation is in the order of 15%. This leads to remarkable variations between the elastic moduli, as summarized in Figure 4, where the specimens are numbered left to right moving along the laminate transversally to the fibers. 1 2 3 4 5 6 7 8 9 0 5 Specimen Figure 4. Variation of E 1 moving along the Biotex/SuperSap laminate. Specimens are numbered from left to right moving transversally to the fibers direction. estimation of the fiber density variation, it can be stated that the variation is in the order of 15%. This leads to remarkable variations between the elastic moduli, as summarized in Figure 4, where the specimens are numbered left to right moving along the laminate transversally to the fibers.

100 125 150 175 200 225 250 275

0 . 2 0 . 4 0 . 6 0 . 8 1 1 . 2 1 . 4

Stress [ MPa ]

25 50 75

Normalized Stress

0 0 . 2 0 . 4 0 . 6 0 . 8 1 1 . 2 1 . 4 1 . 6 1 . 8 2 0

0 0 . 2 0 . 4 0 . 6 0 . 8 1 1 . 2 1 . 4 1 . 6 1 . 8 2 0

Strain [%]

Strain [%]

Figure 5. Left: Original stress-strain curves for a set of BioTex/SuperSap specimens loaded in tension. Right: Results with normalized stress data. Fig. 5. Left: Original stress-strain curves for a set of BioTex / SuperSap specimens loaded in tension. Right: Results with normalized stress data. Notwithstanding the dispersion of the results in terms of elastic moduli, which thus has to be ascribed to varia tions f the fiber volume fraction between specimens, an interesting behavi r is found by norm lizing the t ess. We arb trarily decided to normalize all the stresses by the stress attained a 1.5% strain. Th s way the curves collapse on e sam path, independently of the fiber volume fr ction. It is th refore found that the ratio between E 1 d E 2 constant. Further, the kn e denoting the transition b tw en these two moduli is always l cated at 2000 µε . A thir important finding relat s to the elongation at break. Results s ow a high dispersion in terms f ultimat stress, rang ing between 160 and 250 MPa, but such dispe sion remarkably reduces wh n observing the ultim e strain, which i always attaine in the range 1 . 7 ÷ 1 . 9%. Such result supports the findings by Virk et al. (2010), who propos d the failure strain as key design criterion natural co posites. Notwithstanding the dispersion of the results in terms of elastic moduli, which thus has to be ascribed to variations of the fiber volume fraction between specimens, an interesting behavior is found by normalizing the stress. We arbitrarily decided to normalize all the stresses by the stress attained at 1.5% strain. This way the curves collapse on the same path, independently of the fiber volume fraction. It is therefore found that the ratio between E 1 and E 2 is constant. Further, the knee denoting the transition between these two moduli is always located at 2000 µ Á . A third important finding relates to the elongation at break. Results show a high dispersion in terms of ultimate stress, ranging between 160 and 250 MPa, but such dispersion remarkably reduces when observing the ultimate strain, which is always attained in the range 1 . 7 ÷ 1 . 9 %. Such result supports the findings by Virk et al. (2010), who proposed the failure strain as key design criterion in natural composites. We presented some preliminary results of an experimental campaign on flax composites fabricated through a LRTM process. Results evidenced that the LRTM process can be e ff ectively utilized to manufacture flax composites, obtain ing mechanical properties which are in line with the best properties found in the literature. The behavior of flax composites is found to be both viscoelastic, being highly dependent from the strain rate, and viscoplastic, since inelastic strain is found to accumulate when stains exceed 0 . 1%. The inelastic strains are never recovered and make the apparent modulus of the material to increase at every increasing load cycle. At low strain rates, creep is found to further influence the inelastic strain accumulation. 4. Conclusions