This manuscript presents the study of the microstructural evolution, plastic anisotropy, and mechanical behavior of a duplex stainless steel (DSS) processed by the equal channel angular pressing (ECAP) technique. The ECAP process produced shear bands affecting both phases, austenite and ferrite, which in turns act as preferential sites for the appearance of the new ultrafine grains. Microstructural observations indicated grain sizes smaller than 300 nm in both phases. However, marked differences in the grain boundary misorientations were observed. Most ferrite grain boundaries showed low misorientations (average misorientation of 30°). In contrast, the austenite grain boundaries were mainly dominated by high-angle grain boundaries (average misorientation of 39°). The ECAP processing allowed to reach a yield strength over 1.1 GPa after one ECAP pass. Dislocations formed walls in the ferrite, while they were distributed evenly in the austenite grains creating plastic gradients between the two phases. Through the visco-plastic self-consistent model, it was found that austenite and ferrite strain hardening at different rates, generating plastic instabilities at different strain magnitudes. In this way, it was shown that austenite is the phase that provides more hardening while ferrite provides ductility. Regarding the anisotropy of the steel, crystal plasticity simulations showed that during the first passes of ECAP, the Lankford coefficients increase notably due to the heterogeneous microstructure of sheared grains with a higher density of defects forming subgrains in ferrite than austenite. Moreover, the austenite was more responsible for the larger planar anisotropy ( $$\Delta r$$Δ r = 2.18) values than ferrite ( $$\Delta r$$Δ r = 1.67) after two ECAP passes.

The thermal stability of ARMCO iron processed by equal-channel angular pressing (ECAP) up to a true strain of 16 was investigated by differential scanning calorimetry (DSC). Particularly, the analysis was focused on deriving the recrystallization behavior (onset, peak, and offset temperatures) at three different heating rates (10, 20, and 40 °C/min). Additionally, the stored and activation energies were calculated. As well, the microstructure and the tensile response were evaluated at different numbers of passes before and after annealing heat treatment to assess any significant grain growth and its influence on the material ductility. The different energy contributions (dislocations, grain boundaries, and vacancies) were calculated, being verified that the main contribution came from vacancies.