Human Body Models are essential for real-world occupant protection assessment. With the overall purpose to create a robust human body model which is biofidelic in a variety of crash situations, this study aims to evaluate the biofidelity of the SAFER human body model in far-side impacts. The pelvis, torso and the upper and lower extremities of the SAFER human body model were updated. In addition, the shoulder area was updated for improved shoulder belt interaction in far-side impacts. The model was validated using kinematic corridors based on published human subject test data from two far-side impact set-ups, one simplified and one vehiclebased. The simplified far-side set-up included six configurations with different parameter settings, and the vehicle-based included two configurations:​ with and without far-side airbag, respectively.

The updated SAFER HBM was robust and in general the model predicted the published human subject responses (kinematic CORA score > 0.65) for all configurations in both test set-ups. An exception was a 90 degree far-side impact with the D-ring in the forward position, in the simplified set-up. Here the model could not predict the shoulder belt retention, resulting in a low CORA score. Based on the overall results, the model is considered valid to be used for assessment of far-side impact countermeasures.