ABSTRACT
A rollover accident requires different kinds of analysis, and different kinds of inspection techniques, than do crashes into objects or other vehicles. Its severity is usually not expressed in terms of ΔV. Except for crashes involving a quarter turn from its wheels to its side, it does not consist of a single collision. Instead, it entails a whole series of crashes that occur every time the vehicle touches down on the ground. As a whole, the string of crashes in a rollover is a random, chaotic sequence. It is almost completely lacking in repeatability, which is a primary reason why staged rollover tests have never been prescribed as part of vehicle safety standards. The contact velocity or ΔV of each individual impact that occurred is virtually impossible to determine. While a computer simulation would contain such measures as part of its calculations, rollovers of actual vehicles are unpredictable. For example, in one rollover test, a roof rail may impact the ground and be deformed; in a subsequent test under identical conditions with an identical vehicle, the same roof rail may miss the ground by a fraction of an inch and be essentially unscathed.
