The consequences of average curve generation: implications for biomechanics data

Area of Science:

• Biomechanics and Mechanical Engineering
• Vehicle Safety and Impact Dynamics
• Experimental Data Analysis

Background:

• Understanding complex system mechanical responses often involves impact testing and analyzing time-history data.
• Experimental data frequently contains noise and variability from multiple sources, complicating analysis.
• Averaging multiple test results is a common method to reduce random noise but may obscure fundamental response variations.

Purpose of the Study:

• To evaluate the efficacy of using transducer time-histories for developing an underlying mechanical response when variations are fundamental, not just random noise.
• To explore the limitations of traditional averaging methods when dealing with non-random variability in experimental data.
• To assess the applicability of these methods to the development of Anthropomorphic Test Devices (ATDs).

Main Methods:

• Analysis of transducer time-histories from impact tests on complex systems.
• Comparison of data averaging techniques against methods that account for fundamental response variations.
• Evaluation of data from New Car Assessment Program (NCAP) tests as illustrative examples.

Main Results:

• Averaging noisy data may not accurately represent the underlying physics if variations are fundamental.
• Transducer time-histories offer a potential method to discern and model inherent system responses.
• The analysis highlights challenges in developing accurate Anthropomorphic Test Devices (ATDs) based on averaged post-mortem human surrogate data.

Conclusions:

• Transducer time-histories are crucial for understanding fundamental mechanical responses beyond random noise.
• Careful analysis is needed to avoid misinterpreting averaged data, especially in surrogate development.
• This research provides insights for improving the fidelity of Anthropomorphic Test Devices (ATDs) in impact simulations.