Locking of the strain rate effect in Hopkinson bar testing of a mild steel

The flow stress of many metals exhibits a dependence on the strain rate, in the form of a dynamic amplification of the static flow stress which, according to most literature models, saturates at the strain rates typical of Split Hopkinson Tension bar (SHTB), roughly in the range 500-5000 s(-1). Recently it was found that the dynamic amplification in a mild steel quits evolving at the necking onset, becoming insensitive to further strain rate variations, and remains locked at that current value, independently of the very steep and uncontrollable rise of the strain rate occurring beyond such stage. The present paper investigates in detail this locking effect, analysing how the strain rate increases the flow stress and the necking influences such increase, and then outlining a modelling strategy for such interactions, whose details are analysed on the basis of both experimental measurements and finite elements (f.e.) simulations. By mean of this approach, the locking phenomenon is compared to the saturating feature modelled by the previous literature models, highlighting the differences between them and evaluating their compatibility with the postnecking strain histories from experiments. High-speed camera measurements of the shrinking diameter enable to assess the suitability of the true variables (stress, strain and strain rate) for identifying the material response and for validating material models. The locking effect is also checked against experimental results from the literature, so referring to materials exhibiting necking initiation at both early and intermediate plastic strains. Finally, the locking effect is found to limit the strain rates up to which the SHTB experiments can really reflect the dynamic response of materials. This limitation is especially pronounced for early-necking materials.