This episode of the 2018 Just Science podcast series consists of an audio interview with Dr. Ashraf Bastawros from Iowa State University, who discusses how brittle metal fracture mechanics principles can be used with statistical learning tools to produce quantitative results.
Dr. Bastawros' specialized knowledge pertains to the characteristics of metals and the physics of how they behave under the stress that is sufficient to fracture a particular metal tool. The relevance of this science to criminal investigations occurs in cases where a metal tool such as a knife or chisel fractures in the course of a crime, leaving one piece at the crime scene and the other piece retained in the offender's possession. A match would place the offender at the crime scene. Issues discussed in the interview include the types of fractured metal likely to be found at a crime scene (hard, brittle metals vulnerable to fracture). Dr. Bastawros describes the procedure used to match one piece of metal from a fracture with another piece and derive a statistical probability of a correct match. The interview focuses on distinctive characteristics of metals and of a given fracture of a metal. Requirements for visualizing the distinctive features of metal and their fracture patterns are discussed, along with the development of mathematical algorithms necessary to indicate the likelihood of a match between separate parts of a metal fracture.