PSI - Issue 5

Martin Kadlec et al. / Procedia Structural Integrity 5 (2017) 1342–1348 Petr Homola / Structural Integrity Procedia 00 (2017) 000 – 000

1344

3

Fig. 2. (a) Schematic of the tapering for both alternatives; (b) Plate no. 3 after waterjet cutting of specimens; (c) a specimen gripped in jaws.

The alternative B has one transition from 16 to 11 layers (Fig. 1b). The layup can be coded as follows: (0/90, ±45)∫(0/90, ±45)| (0/90, ±45, 0/90)∫ ±45| (±45, 0/90, ±45, 0/90)∫ (±45, 0/90)| (±45, 0/90)∫. The tapered angle for alternative B was 6 ° . Nominal thicknesses of the sections were 4.96 mm for 16 layers, and 3.41 mm for 11 layers. The plates had these transitions on each side making it 16 layers on both the sides and 11 layers in the middle (Fig. 2a). The internal ply-drops were not exactly positioned in columns due to layers movement during lay-up and the cure phase. Up to two millimetres deviations of ply drop positions in longitudinal direction were observed during the tapered cross-section evaluation. Four plates with dimensions 400 x 400 mm were manufactured. Two plates had alternative A (plates no. 2 and 3) and two plates had alternative B (plates no. 1 and 4). Twelve specimens, 370 mm long and 25 mm wide, were extracted from each plate using an abrasive waterjet cutter. Plate 3 after the jet cutting is shown in Fig. 2b. Specimens from plate no. 3 and no. 4 had additional milling of the edges for better edge surface quality.

2.2. Test procedures

The testing was performed at room temperature. For static strength testing, two specimens of plates no. 1 and no. 2 were selected. The testing was performed according to ASTMD3039M standard (ASTM international (2014)). Instron 55R1185 with ±100 kN load cell and flat jaws were used (Fig. 2c). Crosshead displacement rate of 1 mm/min was used. Fatigue testing was performed according to ASTM D3479M (ASTM International (2013)) on IST Hydropuls Sinus 100 kN and Schenck 250 kN with frequencies in range of 0.5 and 15 Hz at R = 0.1.

2.3. Method of modelling

Finite element (FE) models of the tapered sections were created in ABAQUS 6.13.2 FE software. In the model, each ply was meshed by three layers of hexahedral elements with reduced integration scheme (C8D3R type in Abaqus). The volumes with matrix adjacent to ply drops were meshed by wedges (C3D6 type in Abaqus). Dimensions of brick elements were 1 x 1 x 0.1 mm. The width of both models was 25 mm. As each ply was represented in 3D, orthotropic material was used. Parameters of the ply were used according to (Airbus (1998)) as follows: E 11 = E 22 = 85 GPa, E 33 = 10.5 GPa, G 12 = 4.1 GPa, G 13 = G 23 = 3.48 GPa, ν 12 = 0.05 and ν 13 = ν 23 = 0.41. The volume filled by the matrix was assumed as isotropic material with properties used according to specification in datasheet (TenCate (2016)). Young modulus of PPS resin is 3 800 MPa and Poiss on’s ratio is 0.36. Only linear behaviour of whole composite was used. In order to analyse the interface of plies in location of resin rich region, the interface layer was inserted in an extended model. The FE model without interface layers or interactions was not capable to show shear stress between plies.