PSI - Issue 77

Valeria Lemkova et al. / Procedia Structural Integrity 77 (2026) 279–291 Valeria Lemkova and Florian Schaefer / Structural Integrity Procedia 00 (2026) 000–000

283

5

for 2 h for an additional isothermal investigation. Increasing temperature above 500 ◦ C was not considered in situ due to system limitations. From the gathered SEM images the grain size was determined, if the grain size was on par with the possible resolution. The results were validated particularly for the last step by EBSD. The grain morphology near the metal / ceramic interface was examined by Transmission Kikuchi Di ff raction (TKD, using the EBSD system mentioned in subsection 2.3) in the as-produced state.

2.5. Elucidation of Interfacial Strength

The interface stability of selected specimens was checked by micro-beam bending tests. The micro-beams were milled by FIB (FEI Helios) with a trapezoidal cross-section. The specimen surface was grounded and polished down to a final silica polishing step and not prepared by FIB. The FIB cuts were done perpendicular to the sample surface for the side faces of the beam and the backside faces were cut under an incidence angle of 45 ◦ . The pre-form was milled with 21 nA and the current was step-wise reduced down to 500 µ A , always under grazing incidence. The in situ bending tests in the SEM were performed using a nanoindentation device (ASMEC UNAT2) with a tungsten carbide wedge. Additional ex situ nanoindentaion testing (TI 900, Bruker / Hysitron Inc., MN, USA) was performed to further support the results. To reveal the strain distribution nearby the interface, FEM simulations were implemented to the software Abaqus standard / explicit CAE (Dassault Syste`mes). The planar 2D shell static model of CPE6M (quadratic plain strain with reduced integration) elements was solved with a direct full-newton solver. Material behavior was implemented isotropic elastic with the elastic constants given in Tab. 1 and a Poisson’s ratio of 0.17 for SiO 2 , 0.22 for Al 2 O 3 and 0.31 for Ni. The displacement of the matrix was promoted to the embedded particle region by a tie constraint. Special attention was payed to mesh compatibility. One edge of the model (Fig. 8 (a)) was encastred whereas the opposite edge was shifted parallel by a displacement Dirichlet boundary condition under a constant normal pressure. 2.6. Finite Element Simulations The results of the ex situ heat increase tests with isochronal temperature steps are shown in Fig. 1. Starting at about 400 ◦ C the reference sample from the powder route looses already hardness while the MMC samples show an improved hardness stability with an increasing number of rotations and a thermal stability up 450 ◦ C for the Al 2 O 3 and up to 550 ◦ C for the samples with Si 3 N 4 . The initial hardness rises with an increasing number of rotations up to saturation, reached at 10 rotations. As seen in Fig. 1 (b) the hardness levels from the Ni reference after 10 rotations at room temperature and from the Ni with Al 2 O 3 sample after 50 rotations at 300 ◦ C are comparable. 3.2. Microstructural Evolution The evolution of the microstructure during the heat treatment has been revealed by the ASTM E 112 Je ff ries planimetric method. The number of grains inside a circle and intersecting the circle is estimated. The average grain size is then calculated from the area of the circle (regarding the magnification of the image) and the counted number of grains (Fig. 2). SE images from the in situ heat treatments are shown for the Ni matrix with SiO 2 in Fig. 3. The grains grew just slightly up to 500 ◦ C . The reference Ni sample has lost nanocrystallinity and is ultra-fine grained already at 500 ◦ C as revealed also by EBSD. Fig. 4 features exemplary Si 3 N 4 particles in a Ni matrix after 5 rotations at room temperature and after 50 pre rotations at 300 ◦ C . The images give clear evidence for the mechanical contrast and the incompatibility stresses near the interface arising from this mechanical contrast. Cracks emit from the interface in the hardened Ni matrix. After dispersion at 300 ◦ C no cracks are found and the particles are smaller and rounded. Smaller ceramic dispersoids are found everywhere in the matrix. 3. Results 3.1. Thermal Stability