Quantitative Electron Tomography for Accurate Measurement of Precipitates Microstructure Parameters in Al-Cu-Li Alloys

Mechanical theories show that properties of alloys are strongly dependent on the morphological parameters of their strengthening precipitates. However, accurate measurement of precipitates microstructure parameters is still a challenging task. In this article, we develop a quantitative electron tomography method by combining computer vision technology to accurately characterize the three-dimensional microstructure parameters, such as volume fractions, sizes and distributions, of the T-1 and delta'/theta'/delta' precipitates in Al-Cu-Li(-Mg) alloys. Since they have extremely large aspect-ratios in shape and large numbers in density upon formation in the Al matrix, these thin plate-like precipitates are difficult to be characterized quantitatively without the assistance of computer vision technology. It is shown that the property difference between two peak-aged states of the alloy can be well explained with the quantitative precipitate parameters correctly measured. Using these correct precipitate data, we also tested the validity of current mechanical models for projecting the contribution of precipitates to the strengths of the alloy, demonstrating that quantitative relations between strength and microstructure parameters still need to be refined.