PSI - Issue 52

Leonardo Gunawan et al. / Procedia Structural Integrity 52 (2024) 560–569 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

565

6

Fig. 6. Model of water in bucket and its boundary conditions

3.3. Material Properties The material properties of Al 6082 such as modulus of elasticity, Poisson ratio, and density were taken from the general data of the material, namely E=70 GPa,  = 2900 kg/m 3 , and  =0.33, while the plasticity parameters were taken from Yibo et al. (2013) as shown in Fig. 7. The ductile damage properties required to model failure are fracture energy (G C ), fracture strain, stress triaxiality, and strain rate. The G C value could not be determined with certainty from the existing data, therefore simulations were carried out using several G C values, namely 100, 200, 400, and 600 MPa.mm. A G C value that produced a failure similar to the experimental result was then selected. The fracture strain value was 0.1 which was obtained by drawing a line down from the stress peak in the plasticity data with a gradient equal to the modulus of elasticity value in Fig. 7. Stress triaxiality was set using a common value of 0.33. The strain rate value in this modeling was set as 0 because in various simulations carried out the maximum strain rate that occurs was 53.8 / s which is quite small, far below the high strain rate category and has no effect on the stress-strain curve of aluminum. Water properties referred to freshwater properties in general, namely  = 1000 kg/m 3 , and viscosity  = 0.001 Pa.s, and sound propagation rate in water of C w = 1420 m/s. 3.4. Simulation The full model of the experiment can be seen in Fig. 8. To resemble the actual state of the experiment, the influence of gravity was taken into account. To reduce computational time, the specimen was placed just above the water surface and given an initial velocity of 7.67 m/s which is equivalent to being dropped from 3 m above the water surface without initial velocity. To simulate contact between the base panel and water causing failure, a general contact was made between all exterior faces. The simulation was carried out using ABAQUS Explicit for dynamic numerical modeling with a duration of 0.01 seconds. The data taken was the speed on four nodes at the upper corner of the base frame, as many as 500 data evenly divided throughout the simulation duration. Speed data was then differentiated, smoothed, and filtered in ABAQUS to obtain data acceleration. This was done because the acceleration data obtained fluctuates greatly and the trend could not be seen properly. 4. Results 4.1. Damage Fig.9. shows the final damage of the base plate after experiencing impact using Model 1 for G C values of 600, 400, 200 and 100 MPa.mm. A G C value of 600 MPa.mm could not produce a failure on the base panel and a G c value of 400 MPa.mm produced a failure but with a small size. G C values of 200 and 100 resulted in failures of equivalent size to those encountered in experiments. Judging from the resulting form of failure, all forms of failure resemble the shape of the '+' sign as generated by the simulation of Francesconi (2009), which does not correspond to what was