On the Use of Linear Accelerometers in
Six-DOF Laboratory Motion Replication: A Unified Time-Domain Analysis
Dr. Norman Fitz-Coy,
Dr. Michael Hale
Abstract: While the subject of extracting rigid
body motion information from linear accelerometer measurements is not new,
there has been renewed interest in this subject. This renewed interest is
due, in a large part, to the ability to perform multi-axis motion replication
testing in the laboratory. Performing these laboratory tests require that
appropriate drive file profiles be developed for the laboratory actuators in
order to replicate field data that has been recorded from the test
article. However, the synergistic coupling that occurs between
translational and rotational motions in the measurement of linear accelerations
have lead to many publications which debate the minimum number of acceleration
component measurements that are required to capture the information to be
replicated in the laboratory. The current debate focuses on whether the
minimum is six or nine linear acceleration components.
The purpose of this paper is to provide a unified discussion on the use of
linear accelerometer measurements as references in laboratory-based multi-axis
motion replications. This paper provides a thorough analysis of the
subject based in the time domain using theoretical mechanics. Specifics
regarding the minimum number of field acceleration component measurements and
the spatial distribution of these measurement locations on the rigid body are
discussed. Furthermore, the salient differences between motion
replication that preserves both accelerations and velocities and motion
replication which only preserves accelerations are discussed; the implications
of these differences on the acquisition of the field data are also
addressed.
Simulation results are presented to validate the mathematical developments and
to highlight the differences between the motion re-construction techniques
currently found in the literature.
To return to the
Program, use the back button above