Representative volume elements for plasticity and creep measured from high-resolution microscale strain fields

Abstract

Homogenization techniques have been widely used in upscaling microscale numerical results to predict the macroscopic response of materials. Often homogenization is based on the concept of a representative volume element (RVE) and thus, measuring the RVE size and understanding what factors influence it are key elements in successful material modeling. Here, the size of the RVE for an austenitic stainless steel alloy is experimentally determined under, both separately and combined, plasticity and creep loading conditions with varying parameters (namely stress, temperature, and creep hold time). We use a high-resolution optical digital image correlation (DIC) methodology capable of discerning residual strain inhomogeneities at the microstructural level. Furthermore, by combining the strain results from DIC with surface microstructural information from electron backscatter diffraction (EBSD), the localized strains near grain boundaries can be isolated allowing for quantitative observations related to the deformation mechanisms responsible for strain accumulation. Finally, by comparing the results for RVE size and localized normal to shear strain ratios for different combinations of loading parameters, the relationship between grain-boundary sliding and the resulting heterogeneity of the strain field is explored. Cases where grain-boundary sliding was the dominant deformation mechanism (i.e., at elevated temperature) had considerably smaller RVE sizes (from 4 to 6 times the average grain size) when compared to samples where sliding was not as prevalent (around 10 times the average grain size).