PSI - Issue 77

P.D.A. da Silva et al. / Procedia Structural Integrity 77 (2026) 103–110 Silva et al. / Structural Integrity Procedia 00 (2026) 000–000

105

3

Table 2. Properties of the selected adhesives.

Property RTV 106 DP 8005

E [MPa] t n

0 [MPa] t s

0 [MPa] G

IC [N/mm] G IIC [N/mm] ρ [g/cm 3 ]

v

1.6

2.3 6.3

1.97

2.73

5 6

1.07 1.06

-

590

8.4

1.1

0.3

AV138

4890 1742

41

30.2 28.7

0.35 1.68

4.91

1.7 1.5

0.35

XNR6852 E-2

42.9

18

0.4

2.2. Joint geometry and dimensions This section defines the single-lap joints (SLJ) used to validate the numerical model. The SLJ is composed of two adherends with identical dimensions of 120 mm in length ( L T ) , L O =25 mm, t p =2 mm, and t a =0.2 mm. For the tubular joints (Fig. 1), L O =20 mm, with the adherends measuring 60 mm in length. The internal diameter ( d in ) of the inner adherend was defined as 16 mm, and the external diameter ( d out ) of the outer tube was set to 24.4 mm ( t a =0.2 mm and t p =2 mm). The geometrical modifications and characteristic variables are the outer chamfer (Fig. 1 a; variable α ), inner chamfer (Fig. 1 b; variable θ ), and adhesive fillet (Fig. 1 c; variable β ).

Fig. 1. Tubular adhesive joint and geometrical modifications: (a) outer chamfer; (b) inner chamfer; and (c) adhesive fillet.

2.3. Fabrication and impact tests The specimens used in the validation study comprised SLJ with L O =25 mm. High-strength steel was employed, as detailed in Section 2.1. The bonding surfaces were prepared by sandblasting to enhance surface roughness, and then cleaned with acetone. To maintain precise alignment during bonding, the specimens were secured in a custom mould. After curing the adhesive at room temperature, the excess adhesive was removed using a rotary abrasive process. Impact testing was performed using a Rosand ® Instrumented Falling Weight Impact Tester (Model 5 HV). The test involved dropping a weight of predetermined mass. The specimens were placed vertically in the testing machine, and the upper specimen edge was secured to the machine frame and the specimen aligned. The lower specimen edge, which is hit by the weight during testing, was designed to slide along a guide, maintaining accurate alignment throughout the impact event. The transmitted impact load was recorded using a load cell attached to the upper specimen edge. The impact tests employed a mass of 26 kg, delivering an energy of 40 J. 2.4. Models preprocessing A 2D numerical analysis of the impact tests was conducted using Abaqus ® by strain-rate dependent CZM modelling, both for the validation SLJ, and the tubular joints, which followed the same preprocessing principles. To represent the tubular joint, an axisymmetric modelling approach was adopted. The adherends were mostly modelled using 4-node axisymmetric solid elements (CAX4), employing also 3-node axisymmetric solid elements (CAX3) at the sloped regions due to the geometrical modifications, while the adhesive layer was represented by a row of cohesive