Design guide for stress-strain measurements of strain rate sensitive metals using impact experiments and DIC

Abstract

Strain rate, how fast a material is strained, is known to have an effect on the behaviour of metals. Being able to measure the effect of strain rate in a material provides more reliable material data as input for material models. Strain rates up to 10 per second can be tested using a (fast) hydraulic testing machine. Strain rates upwards from 500 per second can be tested using a Split-Hopkinson bar, but for the strain rates in between no such standard method is available. The goal of this thesis is to provide a design guide for a reliable experiment that measures the effect of strain rate, in the range 10-100 per second, on the tensile stress-strain curve of a metal. The test method proposed in this thesis consists of two parts. The first part is a test using a universal testing machine to determine material behaviour at low strain rates of 0.001-10 per second. A regular dogbone specimen with a longer grip section is used for the UTM tests, which provides the material data to design specimens for the second part. The second part is an impact test where a drophead impacts a specimen, causing it to strain. The specimens are U-shaped strips with a dogbone at either side to test material behaviour at higher strain rates of 10-100 per second. For both tests, strains are recorded in the grip and gauge sections by means of a DIC system. The main advantages of the proposed test method are (i) that no sensors are required in the drophead as the load is extracted from strain measurements in the linear elastic grip section, while the gauge section is allowed to deform plastically and (ii) by using DIC, unobtrusive measurements are taken of the strain field in the recorded area. Two analytical models have been developed, one for the universal testing machine tests and one for the impact tests. The analytical models for the UTM tests and the impact tests have been compared to a finite element model of the same specimen. When plastic strain in the gauge section becomes the most dominant component of the strain, both analytical and FE strain curves show good agreement. Numerical simulations of the impact test have been done by means of an explicit, dynamic, non-linear impact simulation using finite element analysis. A parametric study has been done using the FE model to determine the effect of drophead mass, impact velocity and specimen dimensions on the strain rate in the gauge section and the measurement accuracy. Based on the results of this study, a guideline is presented for performing the experiment. In conclusion, a novel test method and corresponding guideline to determine the stress-strain curve of metals at intermediate strain rates in the range of 10-100 per second has been presented and demonstrated by means of numerical simulations. As a future step, a set of experiments should be performed to prove the validity of the proposed test method.