PSI - Issue 77

L.A.S. Maia et al. / Procedia Structural Integrity 77 (2026) 87–94 Maia et al. / Structural Integrity Procedia 00 (2026) 000–000

90 4

in which the PEEQ parameter was used to identify equivalent plastic deformation in the adherends. Boundary conditions were applied to replicate the experimental test setup: the edge of the adherend opposite to the artificial mass was clamped, while the edge adjacent to the mass was constrained in the transverse direction, allowing only longitudinal motion. The impact load was emulated by a velocity of 1.75 m/s applied to the artificial mass. The mesh consisted of quadrilateral elements with a side length of 0.2 mm. Two types of elements were utilized: plane strain CPE4R elements for the adherends and mass, resulting in 13,050 elements, and 4-node COH2D4 cohesive elements for the adhesives, totalling 125 elements. This configuration produced a mesh comprising 13,175 elements overall.

3. Results and discussion 3.1. Experimental validation

SAJ experimental and numerical results by Valente et al. (2019) were considered. P m is presented in Fig. 3 and summarized in Table 3, including numerical P m deviations (Δ). For the AV138, the deviation was within acceptable limits (numerical P m being 6% lower than the benchmark numerical result and 16% lower than the experimental data). For the DP8005, the deviation between the numerical and reference results was minimal (1%), and the experimental P m was 15% lower. For the XNR6852 E-2, the numerical P m showed a 5% deviation compared to the numerical reference and was 16% higher than the experimental reference. Overall, the CZM approach was thus considered a robust tool for design applications, consistently delivering reliable approximations of expected joint performance.

40

30

20

P m [kN]

10

0

Experimental

Numerical reference

Current numerical

AV138

DP8005

XNR6852 E-2

Fig. 3. Comparison of experimental and numerical P m values for various adhesives.

Table 3. Quantitative comparison of experimental and numerical P m [kN] with deviations.

Adhesive type

AV138

DP8005

XNR6852E-2

Numerical benchmark P m (a)

14.43 11.75 13.61

19.93 16.79 19.90

26.92 24.00 28.39 5.46% 16.31%

Experimental P m (b)

Current numerical P m (c)

Δ (c) – (a) Δ (c) – (b)

-5.68% 12.89%

-0.15% 15.60%

3.2. Outer chamfer effect

3.2.1. Stresses τ xy stresses across the adhesive section of the DP8005+AV138 joint at the boundaries between adhesives ( x / L O =0.17; 0.83, being x the horizontal coordinate), as shown in Fig. 4 (a), range from 2.5 to 3.5 times the average shear stress τ avg . The effect of the outer chamfer is visible between the exterior of the adhesive layer and the adhesive boundaries. Compared to the latter, the RTV106+AV138 joint presented higher peak τ xy / τ avg stresses, ranging from 3.6 to 4.5. In the RTV106 region ( x / L O є [0.00, 0.17]; x / L O є [0.83, 1.00]), τ xy stresses were negligible due to its high flexibility. Among the three joints, DP8005+XNR6852E-2 exhibited the lowest τ xy / τ avg , ranging from 1.7 to 2.