The clavicle is the most commonly fractured bone in the human body. The majority of clavicular fractures (75–80 %) occur within the middle third of the bone. In males, the annual incidence was highest under 20 years of age, decreasing in each subsequent cohort until the seventh decade. In females, the incidence was more constant, but relatively frequent in teenagers and the elderly. The common mechanism of clavicle fracture, regardless of site, is a direct blow on the shoulder such as occurs during a fall or road traffic accident [5].

Several studies have investigated the impact forces involved in clavicle fracture. Duprey et al. conducted dynamic axial compression tests until failure on 18 clavicles using a drop tower system (6 kg impactor mass), at four impact speeds between 1 and 2.5 m/s [6]. A schematic of the system is shown (Fig. 12.7). Clavicle fractures occurred at an average velocity of 1.41 ± 0.4 m/s, with an average fracture force of 1.48 ± 0.46 kN and average deflection of 5.4 ± 1.1 mm. Failure force of female clavicles was significantly lower than the male, and at the slowest impact speed (1 m/s) only female clavicles ruptured. Effect from bone mineral density (BMD) on failure force was not detected due to small samples. In a comprehensive imaging study, Harrington et al. investigated the cross-sectional geometric properties of 15 adult clavicles from five females and ten males [7]. Cross-sectional area, anatomic moments of inertia, and porosity along the length of the clavicle were measured. Mechanical properties, including axial, flexural, and torsional rigidities and critical buckling force were derived using beam theory based on the geometric data. The clavicle was found to be weakest in the central third of its length. Analysis of Euler buckling predicted a minimum critical force for buckling during axial loading of approximately two to three body weights for an average adult, estimation in support of findings from an early study [6].

Mechanical response of the clavicle was reported by Kemper et al. [8]. Ten cadaver clavicles were tested in three–point bending, with impact delivered by a servo hydraulic actuator at a velocity of 0.15 m/s. A schematic of the system is shown (Fig. 12.7). The average failure load found was 732 ± 175 Nm and the average failure moment was 28.3 ± 7.8 Nm. Most clavicles fractured in the mid-shaft region in the form of a transverse, oblique, or comminuted-wedge fracture.

Early tolerance studies of the humerus were conducted using three-point bending loading technique and maximum bending moments were reported. In one of the earliest studies conducted in 1859, according to a 1995 citation moments for males and females were 115 and 73 Nm [9, 10]. Greater moments of 151 and 85 Nm were reported for males and females in the1880 study, according to the cited 1995 reference [11]. In a more recent study reported in 1970, the average fracture load for the humerus was determined to be 1,330 N and this was associated with the quasi-static load applied in the antero-posterior (AP) direction [12]. Studies conducted in the 1990s have determined the effect of loading direction (AP versus lateral-medial, ML) and rate (218 and 0.635 mm/s) on the shear forces, AP and lateral bending moments corresponding to the ML and AP loadings [13]. These authors subjected 24 specimens from one female and 13 male PMHS with age ranging from 44 to 75 years, to three-point bending. The overall moments and shear forces for all combinations of variables were 155 ± 44.6 Nm (±denotes standard deviation) and 1.96 ± 0.67 kN. These authors normalized the data to fifth and fiftieth percentile anthropometries using the impulse-momentum approach [14]. The summary of data is shown in Table 12.1. Because of the lack of statistical difference (p > 0.05) between the two loading directions and rates, the authors suggested the following magnitudes to represent the tolerance of the humerus: fiftieth percentile: 2.5 kN and 230 Nm; and fifth percentile: 1.7 kN and 130 Nm (Table 12.1). Fractures transverse to the longitudinal axis were common and this was followed by angled and spiral/irregular fractures.

A later study used 9.48-kg dropping mass, similar to those used in the clavicle studies described above, described earlier, to impact 12 female humerii under the three-point bending configuration at a velocity of 3.6 m/s [15]. The peak failure moment was 154 ± 27 Nm and the normalized peak moment corresponding to the fifth percentile anthropometry was 128 ± 19 Nm [16]. The impactor force is shown (Fig. 12.8, the dashed and dotted curves are not the subject matter of this chapter).

Like the humerus, early human forearm tests were done using quasi-static loading. For example, a 1970 study found that the failure of ulna and radius occurs at 630 and 520 N [12]. Dynamic tests were conducted to determine effects of pronation and supination tolerance and potential application to airbag loading in motor vehicle crashes [18]. Two drop tests were conducted at 4.6 m/s with a 9-kg impacting mass. In the first, the forearm was pronated with the distal third of the ulna oriented towards the impacting mass (simulating ‘natural driving posture’), and in the second, the specimen was fully supinated position with the anterior surface of the wrist facing the mass. The peak failure loads and moments for the supinated and pronated forearms were 2.1 and 1.5 kN, and 115 and 80 Nm, respectively. Ulna fractures preceded radius fractures in both specimens.

In a later study, Pintar et al. evaluated the response of the PMHS forearms at dynamic rates using a material testing device [19]. They subjected 30 forearm specimens (12 female and 18 male) with age ranging from 41 to 89 years to three-point bending from 15 PMHS (Fig. 12.9). Loading velocities were 3.3 and 7.6 m/s in the two groups of equally divided specimens. Thirteen had butterfly fractures of the ulna (10 from male specimens). Out of the 16 comminuted fracture specimens, 11 were from the high rate of loading. There were no statistical (p > 0.05) differences on the fracture location based on the two loading rates and no statistical differences were apparent with regard to the anatomical aspect (p > 0.05). However, peak force and bending moments were significantly (p < 0.05) lower in female than male specimens. These results clearly indicated that the tolerances of female forearms to three-point bending are lower than males (Table 12.2). Furthermore, compared to the quasi-static tests, these results showed a dynamic factor of 1.7 to account for the increased rate of loading. This factor was found to be in-line with similar data for other bones. For example, the dynamic rise factor is reported to be 1.68 and 1.66 for the tibia and femur bones [20, 21]. Based on this research, using the mean moment of 94 Nm for all specimen as the basis, Pintar et al. proposed a tolerance of 47 Nm for a smaller size occupant [19]. Fractures from this study are shown (Fig. 12.10).