Eight frontal rigid barrier impact tests with one car type were carried out on the test facility of the BASt. The tests were performed in accordance with the draft regulation of the Group of Rapporteurs on Crashworthiness (GRCS) of the Economic Commission for Europe concerning occupant protection in the event of head collisions.

The goal of the test series was to determine the variances of the measured values on the vehicle and on the dummies. Measurements on the vehicle revealed coefficients of variation for the individual values at 2 - 7%. In the case of the measured values for the dummies, a distinction is to be made as to whether they were assumed to have impacted against parts of the vehicle interior or not. Without impact the coefficients of variation of the HIC and SI protection criteria were 10-13%. If an interior impact takes place, the corresponding measured values are, of course, higher. For the head accelerations (maximum values and the 3 ms values) of the dummy in the driver's seat, coefficients of variation of 19 and 13% were measured, while a value of 22% was measured for the HIC. For the thorax the corresponding values for acceleration were 45 and 30% and for the SI they were 33%. In the right femur of the dummy in the passenger's seat, there was an impact between the knee and the instrument panel in every test. In this case the coefficient of variation was 28%.

Knowing the variances of crash test results, design limits can be established to achieve that, in standard-tests, dummy measuring values remain within a defined probability below the legal performance criteria limits.

Although the number of tests conducted for statistical analyses is still too low, some design limits are specified in the assumption of a normal distribution of the measuring data.