The EEVC WG17 upper leg test as used in Euro NCAP was reviewed. Previous work revealed shortcomings of the EEVC WG17 test set-up. Recent published accident data show that injuries to the lower extremities by the bonnet leading edge, not including ground impacts, only accounted for 5% of all AIS2+ injuries and 4% of all AIS3+ injuries. Previous work and this data indicate a discrepancy in importance of the upper leg test between Euro NCAP and real-life injury frequencies.

Suggested legform impactor threshold values have so far not been based on human injury risk transferred to impactor values. The implications of the proposed improvements to the test set-up from Snedeker et al. (2005) for Euro NCAP test results have not been assessed. Both the above issues are aimed at in this study. They are important as only with the right targets and evaluation methods, traffic related injuries can be minimized.

Human injury threshold values for femur and pelvis impact were derived from applicable and original PMHS data. Data was scaled to a mid-sized male, survival analysis with Weibull fit was performed with exact femur 3-point bending data, logistic regression with doubly censored pelvis impact data. Legform thresholds were derived using a linear regression between impactor and THUMS values derived form tests conducted by Snedeker et al. (2005). It is assumed that THUMS and upper leg surrogates have a similar response. The implications of the new set-up and thresholds for Euro NCAP test results were assessed for results published 2009 and 2010 using empirical relationships between impact energy, measured force and moment.

Using this approach, the resulting thresholds to be used with the legform were determined to be 7.9- 9.0 kN for the pelvis test and 300-365 Nm for the femur test. These values correspond to 5 and 20% fracture risk, respectively.

With the currently used set-up and limits, the average score for the upper leg test is 22% of the maximum score. With the proposed method and limits, the average score calculated is 70%. With only 30% missing, the score matches better with the accident frequency of bonnet leading edge induced injuries to lower extremities.