Experimental and numerical investigations of mechanical properties and microstructural evolution of copper processed by accumulative skin pass rolling

Fabricating metallic materials with ultrafine or nanoscale microstructure using the severe plastic deformation (SPD) technique has been shown as an effective strategy to significantly improve the overall performance of the materials. Compared with their coarse-grained counterparts, ultrafine-grained and nanostructured metals and alloys exhibit ultra-high strength and hardness, and superior wear and corrosion resistance. However, these improvements are typically achieved at the cost of a drastic decline in the ductility. This prohibits the use of SPD technique for the manufacture of many engineering components. The strength-ductility trade-off dilemma has partially eased in recent years since novel strategies were developed to produce materials with ‘gradient structure’ using surface mechanical treatments, such as the surface mechanical attrition treatment (SMAT) and the surface mechanical grinding treatment (SMGT). Materials with the gradient structure exhibit desirable properties including excellent strength, corrosion and fatigue resistance, without apparently sacrificing the overall ductility. Unfortunately the sizes of the gradient material samples fabricated by the existing methods are limited. To date, large-scale industrial applications have not been available due to a lack of reliable technologies capable of producing such materials in large volumes. Of the traditional metal forming processes, rolling is the most efficient process for large-volume product manufacture due to its continuous operation. This study aims to utilise the accumulative skin pass rolling (ASPR) process based on rolling technology for bulk manufacture of gradient materials. In this study, ASPR has been successfully applied to produce copper sheets with gradient structure. It has been found that ASPR significantly increases the strength of the copper sheet. The technique postpones the necking and fracture in the tensile test and in turn improves the ductility. The ASPR-processed sample exhibits mechanical properties similar to the gradient materials fabricated using existing manufacturing methods. The experimental and numerical investigations demonstrate that ASPR increases local misorientation and the density of geometrically necessary dislocations (GNDs) near the sample surface, which play a role similar to grain refinement for material strengthening. The lower dislocation density in the centre of the sample is responsible for high ductility. Crystal plasticity finite element method (CPFEM) framework is applied to examine the microstructure evolution of the processed sample. The simulations of ASPR process provide a valuable knowledge to understand the formation of gradient microstructure. Tensile tests using in-situ scanning electron microscopy/electron back-scattered diffraction (SEM/EBSD) were conducted, along with numerical modelling, to investigate the deformation mechanisms of annealed copper and ASPR-processed copper. It was found from the in-situ tests that the formation of ingrain subdivisions depends on the grain orientation and the interaction with neighbouring grains. This affects the slip activities and lead to the variation of slips in different regions, facilitating competitive rotation of lattices that form the grain subdivision. It was also found that the high dislocation density in the ASPR-processed sample makes a more significant contribution to improving the yield strength than the grain boundary strengthening. Finite element method (FEM) modelling shows that the improved yield strength in the ASPR-processed copper is due to the heterogeneity in the properties: the surface layer exhibits higher yield strength than the centre region. Visco-plastic self-consistent (VPSC) simulations are also carried out to better understand the evolution in the microstructure during the tensile tests. It was observed that the intermediate grain-matrix interaction scheme in the VPSC model leads to better agreement with experimental data than the Taylor and tangent schemes