PSI - Issue 24

Riccardo Panciroli et al. / Procedia Structural Integrity 24 (2019) 593–600 R. Panciroli and F. Nerilli / Structural Integrity Procedia 00 (2019) 000–000

596

4

3. Morphing structure as a bistable composite panel

The morphing structure has been conceived as a bistable composite plate by exploiting the thermal deformations introduced as the composite cools down after the cure. Under given combinations of panel dimensions, thickness, lamina properties, and thermal cycles, any antisymmetric lamination sequences might show two stable configurations upon cooling. However, we here concentrate on the [0 90] T lamination sequence only as this is the most prone to show bi-stability (alongside with the [45 -45] T sequence). An e-Glass / Epoxy L antisymmetric panel has been produced to tune the numerical model. The panel has been manufactured utilizing a vacuum bagging process and the suggested curing cycle has been modified to a constant temperature of 100 ◦ C for 24 hours. Considering an ambient temperature of 25 ◦ C the resulting cooling temperature is approximately 75 ◦ C. The final panel dimensions are 400x400x1.30mm. The fiber volume fraction is approximately 50%, and the estimated lamina properties which will be later utilized in the numerical model are reported in Table 2. The mean radius of curvature of the panel has been measured in 280mm. The panel has been then cut realizing two Table 2. E-Glass / Epoxy L material properties. All the properties are estimation from the classical rule of mixture, except for E 11 , which has been evaluated experimentally on UD specimens made for the purpose. E 11 E 22 E 33 G 12 G 23 G 13 ν 12 ν 23 ν 13 40 GPa 12.9 GPa 4.5 GPa 4.5 GPa 0.5 GPa 4.5 GPa 0.27 0.4 0.4 smaller square panels with 150mm, and 200mm sides. The smaller panel was found to loose the bi-stability, showing the typical saddle configuration which is attained when the thermal gradient is not su ffi cient to infer the bistable conditions. The larger square panel instead showed the bistable behavior, but with a minimal curvature. The in-plane thermal expansion coe ffi cients of the lamina have been adjusted to fit with these experimental data, and now read α 11 = 1 . 2 · 10 − 5 and α 22 = 6 · 10 − 5 .

∆ T = 40

∆ T = 20

0 1 2

1

0

− 2 − 1 − 6 − 4 − 2 0

− 1

∆ T = 60

∆ T = 80

0

− 4 − 2

∆ T = 120

∆ T = 100

0

− 8 − 6 − 4 − 2 0

− 10 − 5

Fig. 3. E ff ect of the post-curing cooling temperature on the final curvature of the laminate. Out-of-plane displacements are given in mm.