PSI - Issue 68

Ihssane Kididane et al. / Procedia Structural Integrity 68 (2025) 358–364

360

Ihssane Kididane et al. / Structural Integrity Procedia 00 (2024) 000–000

3

span ( γ = 0 . 5)while γ = 0 . 4 was chosen in case of the CMMB tests. In the latter ones, an initial mode I fracture was forced by moving actuator 2 to a specific displacement δ 2 , 0 while actuator 1 kept F 1 = 0 in force control. A varying mode-mixity was then achieved by prescribing ˙ δ 1 = − 0 . 1mm / s. At the same time, two di ff erent scenarios were imple mented for actuator 2. In the first case, δ 2 = δ 2 , 0 was kept constant, ensuring a strictly monotonic increase in external work during the experiment. In the second case, actuator 2 was unloaded by ˙ δ 2 = − 0 . 04mm / s. Such a test control carries the risk of a temporary unloading, but at the same time allows obtaining mode-mixities that are closer to pure mode II than in the first case. In the SLB tests, ˙ δ 1 = 0 . 1mm / s was prescribed while actuator 2 took over the role of a simple support ( δ 2 ≡ 0).

actuator 1 (piston of testing machine)

actuator 2 (electric motor)

linear slides

load cell 1 shaft encoder 1

load cell 2

shaft encoder 2

sample

shaft encoder 4

crack tip

shaft encoder 3 load cell 3

load cell 4

linear slides

Fig. 2. Labeled picture of CMMB test configuration.

Fig. 2 shows a labeled picture of the test setup in a servo-hydraulic testing machine MTS Landmark ® 370.02 (MTS Systems Corporation, Eden Prairie, Minnesota, USA), which acted as actuator 1. Actuator 2 was an electric motor of type GSX40-0601-IMB-SM9-238-SR (Exlar Europe GmbH, Garching, Germany), which was mounted in the testing machine using aluminum profiles. The standard controller of the testing machine (MTS FlexTest ® 40)was used to move both actuators simultaneously. Each of the support locations and load introduction points were equipped with a S-shaped load cell (ATP Messtechnik, Ettenheim, Germany) and an incremental shaft encoder (Baumer Group, Frauenfeld, Switzerland). For further details on the test setup, reference is made to Kididane et al. (2024).

2.3. Evaluation methods

Referring to the nomenclature given in Fig. 1, the (total) J -integral can be calculated from the forces F i acting on the sample, the corresponding rotations θ i at the load introduction points and the adhesive layer width w ,

1 w ( F 1 θ 1 + F 2 θ 2 − F 3 θ 3 − F 4 θ 4 ) .

(1)

J =

If both (C)MMB and SLB configurations are considered as superimposition of DCB and ENF tests, it is common (e.g. Hong and Yoon (1990); Huber et al. (1993); Rigby and Aliabadi (1998); Ji et al. (2012); Sarrado et al. (2015)) to