PSI - Issue 61

Berkehan Tatli et al. / Procedia Structural Integrity 61 (2024) 12–19 B. Tatli et al. / Structural Integrity Procedia 00 (2024) 000–000

17

6

their left surface to prevent rigid body motion, and the right surface is applied with a displacement boundary condition at a constant strain rate of 0 . 001 s − 1 . A relatively high mesh density, with a characteristic element size of 0.6 µ m, is chosen in the simulations in order to capture the phase field crack paths accurately; moreover, the length scale of the phase field model is set to be five times larger than the characteristic element size. Due to the absence of experimental data and the qualitative nature of this study, phase field fracture toughness values are determined through a trial and error method. The values obtained are 0.30 for the ferrite phase and 0.15 for the martensite phase. Simulations are carried out with the implicit solver of ABAQUS, utilizing a staggered solution scheme for the evolution of both the phase field and the displacement field. The true stress-strain curves are obtained through homogenization using the volumetric averaging method. Fig. 1 presents the distribution of the phase field parameter, true stress-strain response, and Von Mises stress contour plots obtained for the first morphology (Morpho1) in 15% martensite volume fraction (vf15) simulations. The results are obtained for two random crystallographic orientation sets (Ori1 and Ori2), and it should be noted that stress con tour plots are generated at the moment when each simulation reaches its respective ultimate stress points. Firstly, it is deduced that the crystallographic orientation variation significantly influences crack initiation and propagation. While martensite / ferrite interfaces act as stress concentrators, solely examining stress concentration regions is insu ffi cient for deducing the crack propagation path. It is essential to consider the loading history and the misorientation between two adjacent grains. Although the stress concentration regions exhibit some similarities in the Von Mises stress con tour plots, the resultant crack paths diverge significantly from each other. Furthermore, crack paths are observed to predominantly follow transgranular paths within the ferrite grains, occasionally traversing through junctions of the martensite islands.

Fig. 2. Phase Field Distribution (Top) and Engineering Stress-Strain Response (Bottom) of VF15-Morpho2-Oriset1 and VF15-Morpho2-Oriset2.

A similar trend is evident in the results of the second morphology (Morpho2) for the 15% volume fraction (vf15) simulations with two random crystallographic orientation sets (Ori1 and Ori2), as shown in Fig. 2. In this example, a change in the crystallographic orientation set of the ferrite phase has a more pronounced impact on the observed stress levels. Similarly, the resultant crack paths in the two simulations display considerable di ff erences depending on the random orientation sets used. When observing the phase field distribution for the second morphology, it becomes evident from the bright red color of the martensite grains that damage tends to accumulate more at the junctions of the slender martensite grains. The disparities observed in ultimate tensile stress and corresponding strain values in Fig. 2 suggest that despite the volume fraction of the second morphology also being 15%, altering the morphology