Achieving ultrahigh fatigue resistance in AlSi10Mg alloy by additive manufacturing.

Since the first discovery of the fatigue phenomenon in the late 1830s, efforts to fight against fatigue failure have continued. Here we report a fatigue resistance phenomenon in nano-TiB2-decorated AlSi10Mg enabled by additive manufacturing. This fatigue resistance mechanism benefits from the three-dimensional dual-phase cellular nanostructure, which acts as a strong volumetric nanocage to prevent localized damage accumulation, thus inhibiting fatigue crack initiation. The intrinsic fatigue strength limit of nano-TiB2-decorated AlSi10Mg was proven to be close to its tensile strength through the in situ fatigue tests of a defect-free microsample. To demonstrate the practical applicability of this mechanism, printed bulk nano-TiB2-decorated AlSi10Mg achieved fatigue resistance more than double those of other additive manufacturing Al alloys and surpassed those of high-strength wrought Al alloys. This strategy of additive-manufacturing-assisted nanostructure engineering can be extended to the development of other dual-phase fatigue-resistant metals.

Residual-based pattern center calibration in high-resolution electron backscatter diffraction.

High resolution electron backscatter diffraction is an emerging technique of micro-structural characterization which can be used for local elastic strain measurement. Pattern center (PC) coordinate, an important parameter which affects accuracy of HR-EBSD, should be carefully calibrated before calculation. An integrated digital image correlation (IDIC) algorithm can extract the deformation gradient tensor and return the residual between reference and targeted images simultaneously. We propose to use the residual value as a criterion to calibrate PC, as an accurate PC value, accompanied with sample tilt parameters, results in slightly lower level of residuals when using simulated diffraction patterns. Though the reduction of residual value is small in the calibration process, our experimental dataset shows that the calibrated PC value will reduce the retrieved Von Mises strain, which results from the reduction of phantom strain caused by errors in the initially-guessed PC values given by the commercial software DynamicS.