ICF13B

13th International Conference on Fracture June 16–21, 2013, Beijing, China -4- during the creep test an effective coating thickness (ECT) is defined, which is the total thickness of OZ, IDZ and SRZ. The inward diffusion of Al in the substrate is pre-assumed to control the movement of coating-substrate interface in both NiAl and PtAl coatings. The statistic results of the ECT in Fig. 4 (b) presents a linear correlation with √ × , showing as ×√ × (1) In Eq. (1) is the original effective coating thickness and C is a constant. is the diffusion coefficient of Al in the substrate (IN792), calculated by using lower Hashin-Shrtikman bound [13]: ⁄ (2) In Eq. (2) and are the diffusion coefficient of Al in γ and γꞌ phases, and are the volume fraction of γ and γꞌ phases. The values of , , and (listed in Fig. 4(b)) can be calculated by the software DICTRA [14] with Ni-based thermodynamic and kinetic database – TCNI5 and MOBNI2 [15]. The lower Hashin-Shrtikman bound [13] claims the easier-diffusion (high diffusivity) phase as the matrix phase which is likely the case in IN792 with the solid solution γ as the easier-diffusion and matrix phase. Actually applying another Hashin-Shrtikman bound (i.e. upper bound), which assumes the tougher-diffusion (lower diffusivity) phase as the matrix phase, gives the similar calculated results on diffusivity of Al in IN792 [12]. The paralleled curves in Fig. 4(b) give a coating-independent constant C (~ 2.54×105) in Eq. (1) for both NiAl and PtAl diffusion coatings, indicating that the coating thickening rate is controlled only by Al inward diffusion in the substrate. However this conclusion needs to be further checked since the statistic data scattering still exists in Fig. 4(b), especially when the √ × value is lower. Figure 4. (a) Micro-hardness profiles (Vickers) in NiAl and PtAl measured at room temperature (in the same sample in Fig. 3) and (b) Effective Coating Thickness (ECT) near fracture surface vs. √ × . The standard deviations (error bar) in both graphs are given by measurement at ~10 regions of the sample. 3.2. Morphology study of creep fracture Because of heavy oxidation of the fracture surface fractographic analyses cannot be easily applied to study the creep rupture process. Instead, detailed microstructural examinations were carried out on polished lengthwise cross-sections cutting through the mid-width of the creep fractured specimens. The loading direction is in the length direction of the sample and is vertical in all the micrographs shown in this section.

RkJQdWJsaXNoZXIy MjM0NDE=