The ability of three mathematical human body models to predict previously published human responses in two different side impact loading configurations was evaluated using an objective rating method. In particular the kinematics of the shoulder, T1 and head were evaluated. The human body models evaluated were THUMS, HUMOS 2 and the GM model.
The impact loading configurations used were pendulum impact tests and sled tests. In the pendulum configurations, the closest correlation to the published responses was shown by THUMS followed by the GM model. In the sled configuration, closest correlation to the published responses was shown by HUMOS 2 followed by THUMS.
According to the objective rating method the published responses in the pendulum configuration were predicted by all human body models. The published responses in the sled configuration were predicted by HUMOS 2 and THUMS.