Injuries of the upper limbs focus on fractures of the long bones. Of course, the soft tissues and the muscles may also be injured (e.g. skin abrasions due to airbag contact (Duma et al. 2003; Rath et al. 2005)), but these injuries play a minor role in the field of automotive accidents.

As for fractures, the classification presented in Chap.​ 7 also applies. Most common are clavicula fractures which occur, for instance, in direct blows, through compression during lateral impact of the shoulder or in falls on the outstretched arm. A typical fracture of the ulna is the so-called nightstick fracture which is a diaphyseal ulna fracture unaccompanied by a radius fracture. It results from low-energy direct impact (e.g. caused by an airbag) and is characterised by a transverse fracture through the ulna (see Fig.​ 7.​8). Humerus fractures result mainly from direct impact, but can also occur without any contact. Some cases are reported in which muscle forces as involved in overhand throwing caused the humerus to fracture (Levine 2002).

Surveying vehicle crashes in the UK, Frampton et al. (1997) analysed upper extremity injuries of car occupants whose vehicle were not equipped with airbags. It was found that 86 % of all upper extremity injuries were at AIS1 level (minor abrasions, contusions, lacerations). Hence, 14 % formed AIS2+ injuries of which most injuries were fractures whatever the collision type. In frontal collision, forearm fractures were observed most frequently. Shoulder injuries were mainly found in struck-side crashes and rollovers. Clavicle fractures were identified to be the most frequent shoulder injuries. Humerus fractures were found in struck-side crashes but were not common in frontal and rollover crashes. Hand injuries were recorded in some frontal collisions.

Investigating a sample of 540 crashes where the driver airbag deployed, Huelke et al. (1997) found a total of 34 % of drivers sustained AIS1 upper extremity injuries and 3 % sustained AIS2+ injuries to the upper limb.

Kuppa et al. (1997) found an increase from 1.1 to 4.4 % in the occurrence of upper extremity injuries of severity AIS2+ as a result of airbag deployment. A study by Goldman et al. (2005) suggests that injuries to the upper extremity might become more common as a result of an increasing portion of the vehicle fleet being equipped with airbags. In contrast, Segui-Gomez and Baker (2002) who compared vehicles from model years 1993–1997 to vehicles from model year 1998–2001, noted a reduction of upper limb injuries in frontal crashes since the introduction of depowered airbags.

Analysing the US Crash Injury Research Engineering Network (CIREN) database, Conroy et al. (2007) found that the injury pattern differs for drivers and passengers. Only 24.8 % of all occupants had upper extremity injuries. One-half of the injuries to drivers were forearm fractures compared to one-third for passengers. Occupants in side impacts were more likely to have clavicle fractures (29.5 % for passengers vs. 17.1 % for drivers). Airbags were more likely to be a source of forearm fractures, but only 10 % of driver arm fractures with airbag deployment in frontal impacts were associated with airbag fling.

Airbag induced upper extremity injuries in side impacts were also noted by McGwin et al. 2008. Although they did not find an association between side airbag availability and the overall risk of upper extremity, an increased risk for dislocation and AIS2+ injury was observed. In the risk of fracture, however, there was no difference.

In summary, the following causes for upper extremity injuries were identified in the studies mentioned above:

Furthermore, it was observed that clavicula fractures may be caused by the seat belt diagonal section lying across the outboard shoulder and thus transmitting the belt loads transversely across the clavicula.

Additional studies indicate that women are at higher risk in sustaining a AIS2+ upper extremity injury (e.g. Bass et al. 1997; Schneider et al. 1998; Atkinson et al. 2002). It is hypothesised that this is caused by the following factors: (1) women have generally smaller bones resulting in lower ultimate bone strength, (2) women experience an age-related loss of bone mineral density, (3) women are generally shorter in stature and therefore sit closer to the airbag system incorporated in the steering wheel, (4) young adult women tend to add bone to the endocortical surface, in contrast to men who add bone to the periosteal side, resulting in a lower resistance to bending (Schoenau 2001).

Finally, it should be noted that the occurrence of airbag induced upper extremity injuries depends, of course, on the characteristics of the airbag. The term “aggressiveness” is used to describe the influence of airbag design related parameters such as module (cover) design, pressure–time history, seam design and bag folding pattern. The aggressiveness is determined on a relative basis to assess the injury risk between different systems using devices like the Research Arm Injury Device (see Sect. 8.4).

Early work by Weber (1859) and Messerer (1880) determined the load and moment required to produce failure in the bones of the human upper extremities. These studies remained the major reference data until upper extremity injuries received more attention again in the late 1990s. Several research groups addressed the biomechanical response of upper limbs gaining additional data by performing further impact testing. Table 8.1 summarises tolerance values for the humerus reported in the literature.

Concerning forearm fractures, Bass et al. (1997) performed cadaver tests in which ulna nightstick fractures and multiple fractures were observed. The results suggest that the humerus position, the forearm pronation angle and the forearm position relative to the airbag module affect the risk of injury from airbag deployment. Furthermore, it was concluded that there is a forearm strength above which the risk of injury is low, even if the forearm is positioned in front of the airbag module. These findings support the hypotheses that women are at higher risk of sustaining upper extremity injuries.

Investigating the human forearm under a dynamic bending mode, Pintar et al. (1998) determined that the mean failure bending moment for all (male and female) specimens was 94 Nm. However, the bending tolerance of the forearm was found to be highly correlated to bone mineral density, bone area and forearm mass. Consequently, the study suggests that any occupant with lower bone mineral density and lower forearm mass is at higher risk to sustain a fracture.

Cadaver tests by Duma et al. (1998b) addressing the influence of the impact direction showed the forearm to be 21 % stronger in supinated position (91 Nm) than in a pronated position (75 Nm). Conducting additional tests with female forearms in the pronated position and scaling those results to match the 5th percentile female geometry, a tolerance value of 58 Nm was obtained. Given that the forearm is typically pronated in the driving position, the value obtained from this weaker pronation position is meant to represent a conservative injury threshold.

The difference between static and dynamic impact was analysed by Begeman et al. (1999). Bending tests of the forearm were performed both quasi-static and dynamic by using a drop weight which resulted in a loading rate of approximately 3 m/s. Fracture of the ulna or the radius occurred with an average dynamic peak force of 1370 N and an average moment of 89 Nm. Static fracture loads and moments were approximately 20 % lower. Nightstick or simple fractures were the most common type of failure. Differences between the radius and the ulna were not significant. In contrast to the work by Pintar et al. (1998) a correlation of the failure moment with age, cross-sectional properties, bone mineral content or moment of inertia was not found. As tests with one broken bone still showed a high failure moment, the authors suggest that other tissues may play a significant supportive role.

Regarding the elbow, Duma et al. (2001) observed that elbow injuries are caused not only by an axial force but also by a force that acts vertically relative to the horizontal forearm. Hence, a linear combination of the elbow axial and shear force showed a significant correlation to elbow injuries. Carrying out further cadaver tests, Duma et al. (2002) predicted, for the 5th percentile female, a 50 % risk of elbow fracture at a compressive load of 1780 N at an elbow angle of 30° superior to the longitudinal axis of the forearm.

Concerning the shoulder complex, several experimental studies using post mortem human subjects or volunteers investigated the response due to mechanical loading (e.g. Bolte et al. 2003; Compigne et al. 2004) as well as the mechanical properties of the entire shoulder complex (e.g. range of motion and stiffness, Davidsson 2013) or individual anatomical structures (e.g. the shoulder ligaments, Koh et al. 2004). Lateral and oblique impacts as observed in side or frontal-oblique collisions were the focus of these studies. The results reported provide important data, particularly with respect to the definition of response corridors for evaluation and improvement of the biofidelity of current anthropometric test devices. Furthermore, injury threshold values (especially for clavicle fractures) are given, but due to the complexity of the shoulder region the results are not yet conclusive.

To date neither governmental bodies nor consumer test organisations have released guidelines or regulations on how the risk of sustaining upper limb injuries in automotive accidents is to be assessed. There are no conclusive injury criteria or test porticoes implemented yet.

However, Hardy et al. (1997, 2001) presented the concept of the Average Distal Forearm Speed (ADFS) to assess the risk of forearm fractures. Based on static and dynamics airbag deployment tests using cadavers as well as other test devices (i.e. Hybrid III dummy, RAID, SAE arm, for details see below), it was concluded that the distal speed of the forearm which is a function of both the forearm mass and the forearm proximity to the airbag module is a good predictor of the likelihood of forearm fracture. Scaling the results measured to the 5th percentile female geometry, it was found that an ADFS value of 10.5 m/s corresponds to the 50 % probability of fracture. Furthermore, the ADFS decreased linearly with increased distance from the airbag module.

Further studies investigated the possibilities on how upper extremity injuries can be considered in the evaluation of crash tests and on how the aggressiveness of airbags with respect to upper limb injuries may be assessed.

To determine potential forearm fractures due to static deployment of driver airbags, Saul et al. (1996) found that the measurement of meaningful acceleration and bending moments was feasible using a specially instrumented 50th percentile male Hybrid III dummy arm. The arm correlated with airbag aggressiveness, proximity to the airbag module and relative position of the arm with respect to the airbag module.