U.S. flag

An official website of the United States government, Department of Justice.

Physics and Statistical Models for Physical Match Analysis Utilizing 3D Microscopy of Fracture Surfaces

NCJ Number
307872
Date Published
October 2023
Length
60 pages
Annotation

This study uses 3D microscopy of fracture surfaces for physical match analysis via physics and statistical models.

Abstract

This study addresses the scientific and quantitative paradigms for forensic comparative analysis of fractured and torn metal and plastic objects, utilizing three-dimensional (3D) digital representations of their fractured surfaces and their replicas. The research focuses on identifying the domain of unique individuality through spectral analysis of the fracture surface topography and provide a quantitative analysis for match probability and the corresponding error rate that is required to be reported. Three major accomplishments have been reached: Researchers (i) developed mathematical and algorithm framework to identify imaging scales for different classes of materials with intrinsic length scales (metal grain size) and without intrinsic length scale (amorphous plastics), identifying key variables, probability distributions, and quantification of different sources of variability; (ii) performed a systematic study to different fracture scenarios under different loading scenarios (bending, twisting and combination) to achieve realistic forensic data set for fracture comparison; and (iii) performed a validation study on a fully documented consecutively fractured hacksaw sample set provided by our forensic scientist collaborator and utilized for the validation study of fracture match using fracture surfaces topology. Identified key issues of 3D microscopies of surface replicas. The methodology utilized 3D spectral analysis of the fracture surface topography, mapped by 3D microscopy (developed under previous NIJ funding). The framework focused on quantitative statistical measures for the full range of the subclass and individual characteristics of the examined object, and identification of material role on the fracture-feature characteristic scales. Successful advancement of the proposed technique has the potential to provide a new investigative machine-based analysis with quantified error probabilities that can be applied in performing physical matches for a variety of materials. This research will be conducted in response to the NIJ’s expressed need for expanding knowledge underlying forensic science, and in collaboration with forensic scientists working in forensic laboratories.

Date Published: October 1, 2023