PSI - Issue 47

L.A.R. Gomes et al. / Procedia Structural Integrity 47 (2023) 94–101 Gomes et al. / Structural Integrity Procedia 00 (2019) 000 – 000

100

7

the P m vs. L O plots obtained for the SLJ bonded with both adhesives, whose data were collected from the corresponding P -  curves. There is a clear P m increase over the entire L O range for both adhesives, which inclusively is nearly linear for the SLJ bonded with the 7752. This behaviour is linked to the higher flexibility of this adhesive, reflected on the elastic stress distributions (Fig. 3 and Fig. 4), but also its ductility. Considering the AV138 data, P m increases from 6558.9 N ( L O =12.5 mm) to 10440.0 N ( L O =50 mm), although the improvement between L O =25 and 50 mm is not significant. Since this adhesive is compliant (reflecting on the higher peak stress of Fig. 3 and Fig. 4) and brittle, it is expected that the joint performance is affected for large L O due to higher peak stresses and the respective gradients. The SLJ bonded with the 7752 resulted in limit P m of 2366.2 N ( L O =12.5 mm) and 8135.9 N ( L O =50 mm). However, using this adhesive, the P m - L O relationship is nearly linear for the tested range of L O . A relative P m value analysis to the data shows that, for the AV138, the rate of improvement over L O =12.5 mm is either 44.8 or 59.2% for L O =25 and 50 mm, respectively. For the 7752, these quantities increase to 86.8 and 243.8%, by the same order. Actually, for 100 and 300% a proportional relationship would be found on account of the established L O ratios between SLJ geometries. Despite this fact, operating a SLJ with the selected geometrical and material conditions, up to L O =50 mm, the AV138 still manages the best performance between adhesives, although it can be anticipated that, at a given L O , the 7752 will provide better results.

12

12

10

10

8

8

6

6

P [N]

P [N]

4

4

2

2

0

0

0

0.2

0.4

0.6

0.8

1

1.2

0

0.2

0.4

0.6

0.8

1

1.2

δ [mm] 12.5 25 50

δ [mm] 12.5 25 50

a)

b)

Fig. 5. Numerical P -  curves for different L O for the SLJ bonded with the AV138 (a) and 7752 (b).

12

10

8

6

P m [kN]

4

2

0

0

12.5

25

37.5

50

L O [mm] AV138 7752

Fig. 6. P m vs. L O plots for the SLJ bonded with both adhesives.

4. Conclusions The present work aimed at performing a parametric study on the adhesive type and L O effect on the impact strength of composite DLJ. Initial validation was undertaken considering SLJ with similar geometry and materials, serving to corroborate the chosen numerical technique and material properties. Validation was positively accomplished for the AV138, with a P m experimental/numerical relative deviation of 15.8%, which was found acceptable and enabled the numerical study on DLJ that followed. Differences in  P m were higher, but these were attributed to experimental