PAPER ANTHROPOLOGY Mariyam I. Isa, 1 M.A.; Todd W. Fenton, 1 Ph.D.; Trevor Deland, 2 M.S.; and Roger C. Haut, 2 Ph.D. Assessing Impact Direction in 3-point Bending of Human Femora: Incomplete Butterfly Fractures and Fracture Surfaces* ,†,‡ ABSTRACT: Current literature associates bending failure with butterfly fracture, in which fracture initiates transversely at the tensile surface of a bent bone and branches as it propagates toward the impact surface. The orientation of the resulting wedge fragment is often considered diagnostic of impact direction. However, experimental studies indicate bending does not always produce complete butterfly fractures or pro- duces wedge fragments variably in tension or compression, precluding their use in interpreting directionality. This study reports results of experimental 3-point bending tests on thirteen unembalmed human femora. Complete fracture patterns varied following bending failure, but incomplete fractures and fracture surface characteristics were observed in all impacted specimens. A flat, billowy fracture surface was observed in tension, while jagged, angular peaks were observed in compression. Impact direction was accurately reconstructed using incomplete tension wedge butterfly fractures and tension and compression fracture surface criteria in all thirteen specimens. KEYWORDS: forensic science, forensic anthropology, trauma analysis, butterfly fracture, 3-point bending, incomplete fractures, fracture surfaces Direction of blunt force impact is often a factor of interest in forensic death investigations. In the absence of soft tissue evi- dence, forensic anthropologists may be asked to assist forensic pathologists by assessing skeletal fractures and interpreting injury mechanisms. Anthropologists draw on basic principles of biomechanics and bone strength asymmetry to analyze blunt injuries in terms of failure in tension and compression. From this understanding, it is possible to infer the direction of impact (1,2). One pattern of injury often discussed in these terms is the butterfly fracture (1–4). Biomechanical and anthropological liter- ature associates butterfly fracture with long bone bending (3,5–7). When a bone is bent, maximum compressive stress is generated on the concave impact surface and maximum tensile stress on the opposing convex surface. Because cortical bone is weaker in tension than compression, failure initiates in tension opposite the impact (5). As the initial fracture crack approaches the compressed side of the neutral axis, it has been suggested that shear stresses exceed the bone’s shear strength and cause the crack to bifurcate along 45-degree planes of maximum shear (8). The expected result of bending failure is thus a “butterfly” fracture consisting of a transverse segment on the tension side of the neutral axis and a triangular wedge fragment formed by the propagation of oblique fractures on the compression (impact) side (3,4,9). This pattern of fracture is also known as a “tension wedge” because the wedge forms with the apex toward the side of the bone in tension (10). Based on this understanding of but- terfly fracture production, wedge orientation is often regarded as diagnostic of impact direction (3,11–14). Research validating this understanding of butterfly fracture suggests that the relationship between long bone bending and butterfly fractures is more variable than typically presented in the literature (10,15–20). In a key engineering study, Kress (10) reports several complete fracture types generated in bending in a large sample of human femora and tibiae. Tension wedge butter- fly fractures and oblique fractures were most prevalent, but bending also generated transverse and comminuted fractures. Additionally, some butterfly fractures occurred in a “compres- sion wedge” orientation with the apex pointing toward the com- pression side of the bone. Similarly, in an actualistic study of femur and tibia fractures in pedestrian–vehicular impacts, Teresinski and Madro (15) report butterfly fractures occurring in both tension and compression wedge orientations. Other experimental studies provide different reports on the expected prevalence of complete fracture types following bend- ing failure and on the circumstances surrounding the production of compression wedge butterfly fractures (16–20). Martens et al. (16) report 4-point bending of posterior-loaded human femora produced only oblique or compression wedge butterfly fractures. Fenton et al. (17) report no complete butterfly fractures of either wedge orientation in 3-point bending of anterior- and posterior- loaded dry human femora. In the Fenton et al. study, complete 1 Department of Anthropology, Michigan State University, 354 Baker Hall, East Lansing, MI 48824. 2 Orthopaedic Biomechanics Laboratories, Michigan State University, A407 East Fee HAll, East Lansing, MI 48824. *Presented in part at the 67th Annual Meeting of the American Academy of Forensic Sciences, February 16–21, 2015, in Orlando, FL. † Supported by the National Science Foundation Graduate Research Fel- lowship Program (Grant # DGE1424871). ‡ The opinions, findings, and conclusions expressed in this publication are those of the authors and do not necessarily reflect the views of the National Science Foundation. Received 24 Jan. 2017; and in revised form 20 Mar. 2017; accepted 20 Mar. 2017. 1 © 2017 American Academy of Forensic Sciences J Forensic Sci, 2017 doi: 10.1111/1556-4029.13521 Available online at: onlinelibrary.wiley.com