Recent developments of high strain rate servo-hydraulic systems and high speed video
imaging equipment have made the simple tensile test covering the quasi-static to
dynamic range possible. However, obtaining reliable material data from the raw data
produced from such tests requires a good understanding of the unique set of problems
this testing technique can present.
Tensile tests covering the quasi-static to 760 s-1 strain rate range were performed on
Oxygen Free Electrolytic (OFE) copper, Al 6061-T6 and a Ta-2.5%W alloy. Modified
standard sized tensile specimens were used in all tests and evaluations of the higher
strain rate tests were carried out to understand the specimen dynamics at these high
strain rates. Digital Image Correlation was used to measure strain at the higher strain
rates and was ideal as a non-contact extensometer and could provide an indication if
dynamic equilibrium is maintained throughout the test. The work strongly suggests
that each material and specimen geometry will have its own strain rate threshold at
which stress equilibrium is maintained. Appropriate methods were also necessary in
processing the raw dynamic output to extract meaningful material data from the tests.
Data obtained from the tests were successful in evaluating the materials behaviour
over the quasi-static to dynamic strain rate range. The materials responded in a
typical manner to that expected of their crystal structure and stacking fault energy,
agreeing with results available from open literature. The tests performed in tension
were compared with tests carried out in compression and showed the strain rate
sensitivity in tension did not differ substantially to that in compression.
Three constitutive material models were assessed, the Johnson-Cook (J-C) model was
found to represent the experimental results of the OFE Cu and Al 6061-T6 materials well, but did not give such a good fit to the Ta-2.5%W material. The Zerilli-Armstrong (Z-A) model provided a good fit to Ta-2.5%W but not the OFE Cu and Al 6061-T6 materials. No satisfactory fit was achieved using the Mechanical
Threshold Stress (MTS) model.