PSI - Issue 25

Aleksandr Shalimov et al. / Procedia Structural Integrity 25 (2020) 386–393 Author name / Structural Integrity Procedia 00 (2019) 000–000

391

6

a

b

Fig. 4. Values of damage

22 D field for the tensile loaded RVE with pores volume fraction 50% and (a) RVE dimensions 5x5x5 mm; (b) RVE dimensions 10x10x10 mm

Figs. 5 and 6 shows “displacements – force” and “average strain – average stress” curves for the RVEs with different size and volume fraction loaded in the same direction along the Y axis. The selected load triggers the criterium accomplishment and, subsequently, failure of elements already at the first loading steps. The calculations demonstrate that some small size elements already begun to fail, however, the graphs show that during the initial period the structures demonstrate behavior that is close to linear elastic. As it is displayed at Fig. 5, the non-linearity caused by fracture processes developed when the force reached a value of 2.5N for the RVEs with pores volume fraction 50% and edge size 5 and 10 mm, 8N for the RVE with volume fraction 50% and edge size 15 mm. Linear behavior of structures with lesser pores volume fraction of 25% ended at the loading value of approximately 5N for the RVE with edge size 5 mm, 12N for the RVE with edge size 10 mm and 20N for the RVE with edge size 15 mm. The “average stress – average strain” curves presented at Fig. 6 reveals the rate of damage accumulation and influence of the microstructure on stress and strain fields redistribution during deformation process. Correspondence between these curves for the structures with the same volume fraction signify that randomness of microstructure had not impose significant influence on appearance of stress concentrators within the compared RVEs. On the contrary, difference between the curves would be evidence of presence of the stress concentrators that could affect the average values. Thus, comparison of three RVEs with equal volume fraction of ~25% didn’t expose any difference between these curves. However, for a higher volume fraction of 50%, the larger RVE’s average values differs from the rest two structures. Together with the “displacements-force” dependences, this leads to the fact that this particular RVE demonstrated stiffer behavior than the others. To avoid influence of FE mesh size on these conclusions, the same microstructures were calculated using various parameters of meshing. In general, the obtained results showed that influence of the morphological parameters of the microstructure on the fracture behavior of bicontinuous RVEs can be not straightforward in some cases and requires additional studies with more sophisticated mechanical models, particularly, involving investigation of plasticity stage of deformation of ligaments before the fracture.