A common understanding is that in a frontal crash an early coupling of the occupant to the vehicle deceleration is required. This is provided by pretensioning of the belt system. The objective of our study was to set up a rating criterion for pretensioner performance, to benchmark current pretensioner systems, to define requirements for an optimal pretensioning, and to quantify the benefits in both US- and EuroNCAP testing.

A generic test environment was developed and sled tests with different pretensioners and combinations of pretensioners were conducted. As a result, systems with either both retractor and anchor plate pretensioning, or buckle and anchor plate pretensioning gave direct reduction of the dummy chest acceleration values. Additional to the reduction in dummy loading, a reduction in dummy forward displacement occurred. Using this additional space by reduction of the load limiter level of the seat belt resulted in a further reduction in occupant loading, especially in chest deflection. For the determination of the appropriate load limiter level, MADYMO simulation was used. In a further step, a rating criterion for pretensioners was defined. It rates the energy difference of the dummy compared to the vehicle during the crash as percentage of the vehicle energy, i.e. a low figure indicates a good coupling.

As a result, with double pretensioning and respectively tuned load limiter level, chest deflection and acceleration in both EuroNCAP and US-NCAP can be reduced by about 20% - 25% compared to single pretensioning. A low energy difference in the pretensioning rating criterion showed a good correlation to the dummy readings.

With the outcome of the study, requirements to an optimal pretensioning are discussed in respect to a good coupling and to possible injuries induced by aggressive pretensioning.