Spatially resolved eigenstrain analysis across the scales: methods, distributions, insights A. M. Korsunsky∗ Trinity College, University of Oxford, Broad St., Oxford OX1 3BH, UK ∗ alexander.korsunsky@trinity.ox.ac.uk Keywords: diffraction, strain tomography, stress classification, eigenstrain Deformation within hierarchically structured materials is characterized by the complexity that is related to the mechanisms, scale, and the tensorial nature of the eigenstrains (inherent strains) that represent ‘material memory’ of prior inelastic processes. Solid mechanics requirements of total strain compatibility and stress equilibrium link eigenstrain to the measurable distributions of elastic strains within the body. In the course of thermal, environmental and deformation processing of materials, eigenstrains may undergo evolution through plastic deformation, phase transformation, etc. Depending on the scale of consideration, eigenstrains may be associated lattice defects and distortions, slip bands, crack tip zones and other microstructural features. Although it may be possible to quantify eigenstrain distributions directly, in most practical cases they need to be deduced from residual elastic strains (r.e.s.) that can be assessed by non-destructive diffraction techniques or by material removal methods, such as hole drilling or sectioning. Digital Image Correlation (DIC) that has gained popularity as a means of experimental mapping of deformation is characterized by scale independence that allows its application to images obtained at different magnification using various microscopy techniques. A particular application of interest to the present topic is the micro-ring core milling (FIB-DIC for short) as a means of residual stress and eigenstrain evaluation. This approach has made it possible to probe residual stresses of Type I, II and III and to determine their statistical distributions in deformed metallic alloy samples for comparison with crystal plasticity FEM simulations [1]. The author will discuss the possible origins of the observation that elastic strains (and stresses) tend to obey gaussian statistical distributions, while plastic strain (eigenstrain) distributions tend to be lognormal [2]. References [1] Everaerts, J., Salvati, E., Uzun, F., Brandt, L.R., Zhang, H.J., Korsunsky, A.M. (2018) Separating macro-(Type I) and micro-(Type II+ III) residual stresses by ring-core FIB-DIC milling and eigenstrain modelling of a plastically bent titanium alloy bar, Acta Materialia, 156, 43-51. [2] Chen, J., Korsunsky, A.M. (2021) Why is local stress statistics normal, and strain lognormal? Materials & Design, 198, 109319. 16
RkJQdWJsaXNoZXIy MjM0NDE=