PSI - Issue 42

Minghua Cao et al. / Procedia Structural Integrity 42 (2022) 777–784 Minghua Cao et al. / Structural Integrity Procedia 00 (2019) 000–000

779

3

Phenomenological and micromechanical modelling schemes are the main approaches to numerically simulate the thermomechanical behaviour of cast irons. To account for microstructural features, the yield surface and hardening parameters were modified based on phenomenological descriptions (Frishmuth and McLaughlin, 1976). In contrast, the micromechanical methodology focuses on direct modelling of the microstructure of a material, based on experimental observations (Andriollo et al., 2019; Yang et al., 2021). Also, the constituents were assumed anisotropic (Andriollo et al., 2016) or elastoplastic (Andriollo et al., 2015). In this research, both graphite and matrix were assumed as homogeneous and isotropic at the microscale. Thermal debonding as the main fracture mechanism still lacks research about thermal deformation and damage under pure thermal loading. Additionally, it was commonly agreed that the morphology of graphite inclusions affected the macroscopic performance of CGI. This study investigates the effects of graphite morphology on thermal deformation as well as the damage behaviour of graphite under pure thermal loading with three-dimensional numerical models. Specifically, the free-surface profile evolution of graphite and metallic matrix and the onset of interfacial damage at the beginning of debonding are studied in this work. 2. Methodology 2.1. Microstructure characterisation of compacted graphite iron A set of 20 CGI micrographs was obtained with scanning electron microscopy (SEM) and analysed using the software ImageJ. After setting the scale, adjusting the threshold, and outlining the shape of graphite particles (ellipse in this work), the acquired results reflected the microstructure characterisation of CGI (Fig. 2) in terms of geometrical features such as the area, perimeter, volume fraction, major and minor axes lengths of graphite particles.

CGI images

Scale setting

Threshold adjustment

Outline shape adjustment

Results of microstructure of graphite

Fig. 2. Procedure of statistical analysis.

The following assumptions were adopted in 3D numerical simulations, based on the obtained 2D micrographs: • nodular graphite particles in 2D corresponded to spheres in 3D; • vermicular graphite was simulated as a regular oblate ellipsoid; • the diameter of spherical graphite was considered equal to the major length of vermicular graphite; • the rectangular matrix in 2D was considered as a cube with equal length, height, and width in 3D. The dimension of spherical graphite in 2D was calculated with the following expressions = � 2 � 2 ,

= 22 ,

(1) (2)