On the Design of Novel Multi-failure Specimens for Ductile Failure Testing

To quantify uncertainty in the failure response of metallic alloys, conventional experiments may not be suitable owing to the lack of significant scatter in stress states that lead to ductile tearing. In contrast, our testing experience indicates with structures containing strategically located cutouts lead to multiple failure paths and display sufficient scatter in the failure response. Accordingly, we describe the design of dog-bone shaped structures with an ensemble of cutouts, so that at least three different failure paths are observed. In conjunction with Digital Image correlation for full-field displacement measurement and numerical computations, we show how multiple failure paths are obtained when these samples are used. Tests on 2-mm thick 6061-T61 aluminum sheets were conducted using a novel guillotine-style fixture that allows the use of high-speed press. The latter allows testing at strain rates as high as 1 1/s. The primary purpose of this article is to make the case for the use of these samples that not only minimize number of tests required to garner data to calibrate stress-based damage models that capture the entire failure envelope of the material, but also display sufficient scatter that allows quantification of uncertainty in the failure response. (paper)