PSI - Issue 12

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

298

3

dry fibers. The perimeter of such inner chamber is surrounded by a channel 20mm large and 0.5mm thick which is connected to the resin inlet and allows the resin to wet the entire perimeter of the laminate at the beginning of the infusion. The flow channel is surrounded by a first silicon gasket, followed by a vacuum channel 100mm width and a second silicon gasket. Such external vacuum channel allows the mould to remain fully sealed during the infusion process. The vacuum is provided either to the vacuum channel and to the inner chamber at the center of the laminate. As mentioned before, the resin inlet is located within the flow channel, while the resin outlet is through the central vacuum hose. Several moulds with varying thickness have been produced through CNC milling depending on the desired final laminate thickness. All the laminates have been produced to contain approximately 50% of fibers volume fraction. All the laminates followed the curing cycles suggested by the resins manufacturers. Specimens have been cut from the laminates using a CNC router equipped with a diamond end mill. All the specimens dimensions followed the ASTM D3039 suggestions and proper tabbing has been applied. Specimens have been subjected either to monotonic and cyclic tensile tests. Preliminary tests on sacrificial flax / epoxy specimens highlighted a negligible influence of the strain rate e ff ect within the strain rate range 10 − 4 ÷ 10 − 2 s − 1 . A noticeable increase of the mechanical properties has been found at strain rates higher than 10 − 1 s − 1 , while the mechanical behavior at very low strain rates (in the order of 10 − 6 s − 1 ) is found to be very di ff erent, with lower elastic moduli and strength, in particular when utilizing the SuperSap resin. Such behavior should be ascribed to a strong creep influence even at ambient temperature, in line with the findings by Poilaˆne et al. (2014). It has been thus chosen to use a strain rate ˙ ε = 10 − 3 s − 1 to neglect the e ff ect of creep from the results, not being on interest in the present work. Such strain rate has been utilized in both monotonic and cyclic tests. Cyclic loadings have been performed by incrementing the maximum load by defined load steps, and maintaining the strain rate constant during both the loading and the unloading cycles. Strain has been measured through an extensometer with 20mm gage length. The mechanical characterizations have been performed on an 50kN electro-mechanic machine by MTS. 2.3. Mechanical tests The first laminate manufactured and tested has been a Flax / Epoxy 1.2 mm thick. Results are very repeatable both in terms of elastic moduli and strength. However, we must comment that during the entire experimental campaign failure located in the nearby of the tabs in almost every specimen. The location of the failures is listed as acceptable in the ASTM 3039, but it might lead to some scatter on the estimated final strength. Further, some tabs failed during the tests and results are hence not reported in Table 2. Results, summarized in Table 2, show that the repeatability of the mechanical properties is very high for all the parameters on interest. In particular, the values for the final strength, whose mean value is 359 MPa, is in line with the best results found in the literature, suggesting that the manufacturing methodology utilized within this study is adequate. Table 2. Results for the Flax / Epoxy specimens. Results not listed refer to specimens where the tabs failed prior to final fracture. Specimen E 1 σ R ε R 1 29.78 353.8 1.63 2 33.21 349.2 1.72 3 29.46 366.0 1.74 4 31.32 348.5 1.51 5 32.09 378.7 1.73 6 30.10 - - 7 28.91 - - 8 28.88 - - 3. Results

mean (st.dev)

30.4 (1.7)

359 (12)

1.66 (0.09)