Using surround DIC to extract true stress-strain curve from uniaxial tension experiments

The exact stress-strain curve can be directly identified from uniaxial tension experiments up to the point of onset of diffuse necking. To gain insight into the post-necking hardening response of metals, other experimental methods such as the bulge, compression or torsion experiments are typically employed. Here, we introduce the idea of using surround DIC to obtain accurate specimen shape measurements of tensile specimens with rectangular gage section. Based on the results from a series of detailed three-dimensional finite element simulations of ASTM E8 type of uniaxial tension experiments on a wide spectrum of steel and aluminum behaviors, it is proposed to combine the history of the axial strain on the surface at the specimen center with the average axial stress to extract the stress-strain curve for strains of up to axial true strain of 1. The estimation uncertainty of this stress-strain curve estimation procedure is of the same order as that of the associated experimental uncertainties. This result is validated through an additional computational study for more than 100 distinct hardening behaviors. Furthermore, a surround DIC system composed of four stereo DIC systems is built and used to determine the stress-strain curve for a 1.5 mm thick DP780 steel. In addition, bulge experiments are performed on the same material revealing a good agreement of the post-necking hardening behavior determined through uniaxial and bulge testing.

Automated summary

The stress-strain curve can be obtained from uniaxial tension experiments until the onset of diffuse necking. Other experimental methods such as bulge, compression or torsion experiments are used to understand the post-necking hardening response of metals. This study proposes using surround DIC to accurately measure the shape of tensile specimens. By combining the axial strain history with the average axial stress, the stress-strain curve up to a strain of 1 can be extracted. The estimation uncertainty of this procedure is similar to experimental uncertainties.