Error analysis of curvature-based contour measurement devices

Curvature-based contour measurement devices with discrete curvature measurement gauges are widely used for the measurement of dynamic thoracic contours in both dummy and cadaveric automobile sled testing. Such devices include the chestband used to determine local thoracic contours at several rib levels for evaluation of injury parameters in dummy and cadaveric subjects. The use of these devices involves integration of local curvatures to obtain position data, and often incorporates several approximations, including a quasi-continuous approximation of discrete measured curvatures. By comparing a reference and a calculated position profile, this study analyzes the error in local positions induced from several sources. The first source of error is the measurement of curvatures at discrete locations, typically with 2.5 - 5.0 cm curvature gauge spacing. Potential differences in integrated position measurements are calculated and compared for a contour measurement device following corrugations of various widths and heights. The second source of error is a limitation in the curvature full scale. Position errors from exceeding typical curvature limits are found. The study provides a comparison of physical contour parameters and full scale limitations against measurement accuracy for a range of parameters typical of existing dynamic tests. This investigation also includes error analysis of contour measurement devices of previous dummy and cadaver tests. Recommendations are made for acceptable parameter ranges for dummy and cadaveric thoracic sled testing and for the design of enhanced thoracic contour measurement devices. Copyright © 2000 Society of Automotive Engineers, Inc.

Duke Scholars

Author Cameron R. Bass Biomedical Engineering