PSI - Issue 42

Jakub Šedek et al. / Procedia Structural Integrity 42 (2022) 398–403 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

401

4

Table 2. Material properties used for analysis according to Cytec(2012) and Tan et al. (2016). C/PEKK unidirectional ply σ 1F tension 2463 MPa σ 1F compression 1493 MPa E 1 tension 139 GPa E 1 compression 127 GPa* σ 2F tension 61 MPa σ 2F compression 254 MPa E 2 tension 10.3 GPa ν 12 0.3 G 12 5.2 GPa τ 12F 81 MPa G 1c 1.6 kJ/m 2 G 2c 2.1 kJ/m 2 Density 1623 kg/m 3 *E 1 = 100 GPa used to reflect the panel stiffness during the test

A finite element model was created for both, pristine and cracked panels. The ABAQUS package was utilized for analyses. FE models were prepared in ABAQUS/CAE using PbPModGen plug- in developed by Šedek at al. (2022) and the CAE was used for postprocessing. The PbPModGen plug-in implements ply-by-ply (PbP) technique and facilitates the build of detailed FE model so that each lamina is composed by a layer of elements. The continuum shell elements are represented by hexahedral elements, but internally the theory of shells is used. There were 390 000 elements in total. The detailed views of the model are shown in Fig. 3. The connection of the type of skin-web and web-cap was realized by cohesive elements utilizing cohesive zone modelling (CZM). The filler was also included as assumed by 3D model. The crack was included by omitting cohesive elements under the centre stringer between the skin and the web. The damage of plies was considered according to Hashin and Rotem (1973). The damage development was also allowed after reaching failure in some sub-criteria. The loss of stiffness after failure was assumed linear. Material parameters used in the failure criteria were obtained from Table 2. Boundary conditions were applied according to the test in the upper and bottom parts of the panel. The support was represented by encastre constraint in the bottom part. The loading driven by the displacement of the upper part was realized by kinematic coupling with the master node in the centre of the cross-section at the top. All degrees of freedom for the master node were removed except the displacement in the longitudinal axis.

K, ^/s > D Ed^

&/>> Z

&/>> Z

^

ĂͿ

ďͿ

Fig. 3: PbP model of the panel: a) cap-web detail, b) web-skin detail

4.1. Analysis results

The detailed model was solved using the explicit solver. The response up to the point of the stiffness change was linear (see Fig. 4a). Nevertheless, Young’s modulus E 1 was decreased to 100 MPa to reflect the stiffness determined during the test. In comparison with the test, the loading curves of pristine and cracked panels were a little lower further.