A New Shoulder for the THOR Dummy Intended for Oblique Collisions
Doctoral thesis, 2008
Offset and oblique frontal car collisions represent a type of crash in which severe and fatal injuries frequently occur. This indicates the importance of having protective systems in vehicles for these types of collisions. Evaluation of such protective systems requires access to crash test dummies that accurately replicate the human body kinematics. The objectives of this research are to develop and evaluate a new shoulder that is anticipated to facilitate the development of safety systems for cars involved in offset and oblique frontal collisions.
First, the performance of current frontal impact dummies in oblique impact situations was investigated. Tests were carried out in far-side (away from the shoulder belt anchor) and near-side (towards the shoulder belt anchor) collisions. The Hybrid III dummy did not show human-like head kinematics. The THOR dummy showed human-like head kinematics, but slipped out of the shoulder belt more easily than a post mortem human subject (PMHS) in the 45° far-side collisions. It was found that the THOR shoulder may not be adequately biofidelic.
Next, five volunteers and a THOR dummy were tested in a rig where both arms were statically loaded in three directions, forward, diagonally forward-upwards and upwards, while the sternum was supported. The displacement between the right shoulder and the sternum was estimated by means of photo analysis. The THOR dummy was observed to be stiffer and to allow a smaller shoulder range-of-motion then the volunteers. Three belted PMHSs and a THOR dummy were then tested in 0° full-frontal tests, in 45° far-side tests and in 30° near-side tests to study the shoulder and belt-to-shoulder interaction. High-speed video recorded the motions of the test subjects. For the 45° far-side impact, it was concluded that the PMHSs did not slip out of the shoulder belt as easily as the THOR NT did. Further analysis revealed that the geometrical properties of the shoulder bones may be important for the interaction with the shoulder belt in oblique impacts. The response of the human shoulder complex may also influence the head kinematics and thereby head-to-door interaction in oblique near-side collisions.
A new mechanical shoulder prototype was developed for the THOR NT dummy, the THOR SD-1NT. Its design was based on results from the volunteer and the first PMHS study. The shoulder design, SD-1, was intended to have bony surface landmarks and a range-of-motion similar to that of a human, to enable improved belt-to-shoulder interaction during a collision. To resemble the human shoulder stiffness during anterior motion, the SD-1 was designed to have a linearly increasing resistance. The THOR SD-1NT was evaluated against the second PMHS study. It is believed that the geometrical properties and bony landmarks of the SD-1 are in accordance with the human anatomy, and that they provide a human-like belt-to-shoulder interaction, especially during oblique impacts. However, the SD-1 needs further tuning, especially the shoulder range-of-motion. Therefore, a second prototype of the shoulder design, SD-2, was built.
Post Mortem Human Subjects
Crash test dummy
Hörsal Beta, hus Saga, Hörselgången 4, Lindholmen
Opponent: Professor Richard Kent, Center of Applied Biomechanics, University of Virginia, USA.