The analysis presented in this paper indicates that while 1700 lbf (7560 N) is a realistic femur fracture load for 30-50 ms duration impacts, the human femur can withstand higher loads for shorter-duration impacts. Experimental femur fracture data from cadaver and bone specimen tests are reviewed. These data are used to develop femur load fracture tolerance as a function of impact duration.

On the basis of a measured 10% amplification of 1-2 ms input forces by the dummy, the cadaver fracture tolerance is proportionately adjusted to arrive at equivalent load levels for forces measured on current dummy test devices.

Experimental dummy test device data are included and compared to the theoretical response of a mathematical model of the human upper leg. This comparison demonstrates that even neglecting the 10% amplification, there are still significant differences in the response of dummy and human upper leg structures for impact durations less than 3 ms.