Direct monitoring of twinning/detwinning in a TWIP steel under reversed cyclic loading

In situ tensile and reversed cyclic tests were run on a TWIP steel in a SEM, with High-Resolution Digital Image Correlation (HR-DIC) measurements of the plastic strain field in a few selected grains prone to twinning, with a spatial resolution between 150 and 250 nm, or under an AFM, with measurements of surface steps height at emerging deformation twins. Evidences of detwinning upon load reversal, as well as quantitative data on twinning/detwinning/retwinning were obtained. Detwinning and retwinning, which often were only partial, in spite of a fully reversed loading, did not seem to start at the onset of stress reversal. It required a sufficient variation of the stress, close to the twinning stress (estimated as 400?475 MPa) in absolute value, so that a mechanical hysteresis of the local twinned fraction occurred. Primary and secondary twinning along the same plane, inducing axial plastic strains in opposite directions, also allowed some grains to accommodate reversed plastic strain. Under fixed stress amplitude (?500 MPa), the twin fraction in all monitored grains saturated at values between 0.5 and 3.5%, from the 2nd cycle, while under fixed plastic strain amplitude (?0.5%), it increased in a ratchetting way during the whole cyclic hardening stage, reaching 0.5?5%. In both cases, however, the plastic strain amplitude accommodated by twinning/detwinning, which reached 0.35?0.42% in some grains during the 1st cycle, decreased down to less than 0.05% after 100 to 1000 cycles.

Automated summary

In situ tensile and reversed cyclic tests were conducted on a TWIP steel in a SEM, showing evidence of twinning/detwinning processes and their effects. It was observed that detwinning and retwinning occurred partially after load reversal, with stress variations close to twinning stress affecting the mechanical hysteresis of the material.