13th International Conference on Fracture June 16–21, 2013, Beijing, China -3- The final heat treatment step was the same for both microstructures consisting of a plain stress relieve annealing above the Ti3Al solvus temperature. The phase fractions are slightly different (αp: 71% ma, 61% sht), and the mean grain sizes are larger for the sht condition (αp: 3.1µm ma, 7.3µm sht; αs: 3.3µm ma, 11.3µm sht). It should be mentioned that no subgrains were accounted for in the ma condition and that the αs size refers to the colonies, meaning lamellae of the same orientation. Of course the mean lamella thickness is much smaller in case of the ma condition. Figure 1. Microstructure of (a) ma condition and (b) sht condition. All specimens except of the micro-sections were electro-chemically polished in a solution of perchloric acid and methanol prior to testing to meet the requirements for electron backscattered diffraction (EBSD) experiments. The surface should be as smooth as possible without any residual stresses from machining. This requirement can be perfectly met by grinding and vibration polishing (which was done in case of flat specimens for micro-sections), but is impossible for cylindrical specimens, where electrolytic polishing is the only possibility available. It should be pointed out that a slight surface roughness results from different metal removal rates of the different phases. The topography, measured by atomic force microscopy (AFM), was about 400nm from peak to valley. Fatigue experiments were carried out in symmetrical push-pull with a servohydraulic MTS 810 test system under load control at a frequency of 20Hz. All experiments were stopped by specimen failure or at 6x106 cycles. The samples which survived this number of cycles are referred to as run-out specimens. Cylindrical specimens with a milled shallow notch in the gauge length were used for the crack initiation and short crack growth investigations in order to limit the area to be observed. Some tests were interrupted after certain numbers of cycles to enable studying several cracks in the scanning electron microscope (SEM) together with EBSD. A Philips XL30 microscope equipped with automated orientation imaging microscopy (OIMTM) was employed for all analytical research work. A FEI Helios DualBeam Microscope was used for sectioning cracks into the depth. 3. Experimental results Two types of crack initiation sites were observed. Firstly, crack initiation is observed at the interface between two lamellae. Because of the Burgers relationship between alpha and beta phase a b
RkJQdWJsaXNoZXIy MjM0NDE=