Normalizing and scaling of data to derive human response corridors from impact tests

It is well known that variability is inherent in any biological experiment. Human cadavers (Post-Mortem Human Subjects, PMHS) are routinely used to determine responses to impact loading for crashworthiness applications including civilian (motor vehicle) and military environments. It is important to transform measured variables from PMHS tests (accelerations, forces and deflections) to a standard or reference population, termed normalization. The transformation process should account for inter-specimen variations with some underlying assumptions used during normalization. Scaling is a process by which normalized responses are converted from one standard to another (example, mid-size adult male to large-male and small-size female adults, and to pediatric populations). These responses are used to derive corridors to assess the biofidelity of anthropomorphic test devices (crash dummies) used to predict injury in impact environments and design injury mitigating devices. This survey examines the pros and cons of different approaches for obtaining normalized and scaled responses and corridors used in biomechanical studies for over four decades. Specifically, the equal-stress equal-velocity and impulse-momentum methods along with their variations are discussed in this review. Methods ranging from subjective to quasi-static loading to different approaches are discussed for deriving temporal mean and plus minus one standard deviation human corridors of time-varying fundamental responses and cross variables (e.g., force-deflection). The survey offers some insights into the potential efficacy of these approaches with examples from recent impact tests and concludes with recommendations for future studies. The importance of considering various parameters during the experimental design of human impact tests is stressed.