13th International Conference on Fracture June 16–21, 2013, Beijing, China -3- 2.2. Mechanical testing 2.2.1. Uniaxial tensile tests At least two full-thickness smooth tensile specimens (64x12 mm² in gauge dimensions) were cut across each welded joint and pulled in tension, in room temperature laboratory air, using a 250 kN servohydraulic testing machine under displacement control (elongation rate: 2.10-4 s-1 based on a gauge length of 14 mm). Local strains were monitored thanks to a random black-and-white speckle deposited onto the specimen and to conventional strain field monitoring techniques. Incremental calculation of displacements and strains was selected. To determine the onset of KB opening, an in situ tensile test on a miniature dogbone specimen (full thickness, 3×13 mm² in gauge dimensions) was carried out in the SEM, while observing the slightly polished bottom surface of the nugget. 2.2.2. Uniaxial fatigue tests Uniaxial fatigue tests were performed on full-thickness specimens along TD (Fig. 3c,d). A straight gauge part was chosen for the cross-weld specimen, in order to study the fatigue crack initiation site under a homogeneous nominal stress (i.e., load divided by the initial area of the bearing section). All specimen corners and edges were slightly ground using 1200-grit SiC paper, except otherwise stated, to limit the influence of surface irregularities on fatigue crack initiation. Fatigue specimens were tested in room temperature laboratory air, under load control, with a load ratio of 0.1 and a frequency of 20 Hz. Most of the tests were carried out up to fracture. A few tests were started at a lower frequency (0.5 Hz) for the first 10 cycles together with strain field monitoring on one edge thanks to the digital image correlation as described previously. Then, the frequency was changed to 20Hz. After a given fraction of the estimated lifetime at that load level, these tests were interrupted and the specimens were fractured under uniaxial tension. 2.3. Microstructural observations Metallographic samples were polished in cross-section with diamond paste and first etched by anodic oxidation (3% aqueous solution of tetrafluoroboric acid in water, under 30 V with respect to pure Al, for 2-3 min) and then further chemically etched with the Dix-Keller reagent (2 mL HF, 3 mL HCl, 20 mL HNO3, and 175 mL distilled water) and observed using optical microscopy under white light. Fracture surfaces were observed using a Leo 1540VP or a Zeiss DSM 982 Gemini scanning electron microscope (SEM) in secondary electron imaging. After each test, one specimen half was cut along the TD-ND plane at the crack initiation site, polished and etched using the above procedure to determine the influence of welding defects on crack initiation and propagation. 3. Experimental results and discussion 3.1. Tensile properties No significant effect of the welding conditions was detected on tensile properties of JLR and KB bearing welds, with respect to sound welds (Fig. 3b). The yield strength and tensile strength of the welds were lower than those of the base metal (along the same loading direction) by about 40 and 20%, respectively (Table 1). Strain field measurements yielded identical results for all welds, except for a slight strain localisation at the bottom surface of the Gap7 bearing nugget [11].
RkJQdWJsaXNoZXIy MjM0NDE=