PSI - Issue 23

Keng Jiang et al. / Procedia Structural Integrity 23 (2019) 451–456

452

Keng Jiang et al./ Structural Integrity Procedia 00 (2019) 000 – 000

2

construction, mining, transportation, and oil industries, the tungsten carbide powder market is forecasted to reach two billion euros by 2023 (P&S Intelligence (2018)). Under most of operating conditions, hardmetal components are subjected to cyclic loading, and thus fatigue fracture is one of the most common types of failure. The study on fatigue behavior of hardmetals becomes popular among materials engineers and scientists. From a materials science point of view, the superior mechanical performance of WC-Co hardmetals benefits from combination of its constituents, i.e., the brittle WC phase contributes the high hardness and strength whereas the ductile Co phase compensates to the loss in toughness. It yields the fact that the fatigue performance of hardmetals strongly depends on its microstructure. Experimental investigations on the fatigue crack growth (FCG) indicated that the fatigue sensitivity of this material is significantly dependent upon binder content or carbide contiguity, while a monotonic correlation was not found between the FCG threshold and above microstructure features (Schleinkofer et al. (1996), Sailer et al. (2001), Llanes et al. (2002)). With the rise of finite element (FE) analysis, numerical method is also applied in investigating the microscopic damage and fracture of hardmetals. Fischmeister et al. (1988), Mishnaevsky et al. (1999), McHugh and Connolly (2003) carried out early efforts, implying that the fracture process in the microstructure of hardmetals can be numerically simulated by modeling the nucleation, growth and coalescence of microvoids in the binder phase. Recently, the numerical method applied by Özden et al. (2016) shows the possibility of simulating microscopic FCG of hardmetals with failure models of both phases for the case of cyclic loading. It is important to notice that most of current numerical work only focused on the failure behavior of the ductile binder phase, while brittle fracture in WC phase is omitted, and only the damage mechanism under monotonic loading was studied. Besides, realistic microstructure has been rarely reflected in created models. From the perspective in crystallography, local crystalline orientation, anisotropy and fracture strength of a single WC grain have never been modeled. Therefore, this study takes into account the morphology as well as local material parameters of single WC crystals. Benefiting from advanced electron backscatter diffraction (EBSD) characterization technique, the study is able to reconstruct more precise microstructure models. Finally yet importantly, the influence of micro- and mesoscopic residual stresses on the fatigue property is also considered for the first time in numerical simulation. Both the morphology and local material parameters must be properly characterized for the microstructure of hardmetals. Transmission electron microscopy (TEM) investigation shows that WC grains usually exist as truncated triangular prisms (Lay et al. (2008)). With scanning electron microscopy (SEM) as Fig. 1 shows, irregular polygons are usually observed as sections of polyhedra. The morphology comes from the nature of hexagonal lattice structure of WC crystals. The values of linear elasticity constants to connect the stress tensor and strain tensor of WC crystals have been determined by Lee and Gilmore (1982), Golovchan (1998). Referring to the microbeam testing introduced by Trueba et al. (2014), in which the fracture strength of individual WC grains was measured by in situ nanoindentation, it is assumed the brittle fracture is the form of failure for WC grains, thus the maximum principal stress strength theory is applied. The material is marked as failure as soon as the value of 1 exceeds its criterion 1,cr , 2. Materials characterization 2.1. WC phase

0, if 1, if D    

1 1, cr 1 1, cr      

(1)

The material parameters for WC phase are summarized in Table 1. Table 1. Material parameters for WC phase (Lee et al. (1982), Golovchan (1998), Trueba et al. (2014)). 11 [MPa] 12 [MPa] 13 [MPa] 33 [MPa] 44 [MPa] 66 [MPa] 1,cr [MPa] 720 000 254 000 150 000 972 000 328 000 233 000 6 300