PSI - Issue 62

Silvia Caprili et al. / Procedia Structural Integrity 62 (2024) 355–360 Caprili et al. / Structural Integrity Procedia 00 (2019) 000 – 000

359

5

3D nonlinear analyses are ongoing on selected strands that were subjected both to mechanical characterization and X Ray tomographies, using ABAQUS software. Preliminary tests were executed, starting from the single wire model and the strands in reference and corroded conditions, with the aim to define some useful parameters such as the optimal size of the mesh and the influence of pits characteristics on response (depth, orientation along the wires, etc.). The strand model includes both king wire and each of the outer wires; a length equal to 80 cm is considered, as the free length of the element during the tensile tests set-up. Each model is ended restrained, with the aim of reproducing the real conditions of the tests: in particular, on one end, a roller is present to allow the longitudinal elongation while the other end is fixed. Fig. 4 shows some details about the restraints and the model of strand in reference conditions.

(a)

(b)

(c)

Fig. 4. Detail of restraints (a , b) and of mesh of FEM Model (c).

In corroded conditions, the pits were explicitly reproduced at geometrical level, according to what already suggested in literature (Stewart, 2008; Jeon et al. 2019): following the results of X-Ray tomographies, a half-ellipsoidal portion of steel is removed, in different positions along the length of the wire, as is described by Fig. 5.

Fig. 5. Details of pits modelling.

To reduce the computational effort, nonlinear Hexaedrale C3D8R element type were adopted in correspondence of pits regions while in the rest of the element, Tetrahedral linear elements were used. In the same way, the sides size of the mesh are different in correspondence of pits zones respect to the other parts, with a dimension equal about to the half of pit depth, and therefore up to 0.1 mm in the first case, while between 0.3 mm and 2.5 mm in the second one. Regarding the constitutive law of the steel, traditional elasto-plastic law was adopted, while Ductile Damage Criteria was used to simulate the failure criterion, calibrated on the results of experimental tests mainly in terms of achieved ultimate deformation. A further aspect is related to the laws adopted to describe the interaction of wires with each other: a ‘surface to surface’ approach was adopted, with a n infinite stiffness in normal direction and a friction law in tangential too. In Fig. 6 a first result obtained from the calibration is presented, allowing to appreciate the good agreement between experimental test results and 3D FEM model simulation .