A split Hopkinson bar technique to evaluate the performance of accelerometers

We developed a split Hopkinson bar technique to evaluate the performance of accelerometers that measure large amplitude pulses. A nondispersive stress pulse propagates in an aluminium bar and interacts with a tungsten or steel disk at the end of the bar, We measure stress at the aluminum bar-disk interface with a quartz gage and measure acceleration at the fi ee end of the disk with an accelerometer The rise time of the incident stress pulse in the aluminum bar is long enough and the disk length is shout Enough that the response of the disk can be approximated closely as rigid-body motion; an experimentally verified analytical model supports this assumption. Since the cross-sectional area and mass of the disk are known, we calculate acceleration of the rigid disk from the stress measurement and Newton's Second Law. Comparisons of accelerations calculated from the quartz gage data and measured acceleration data show excellent agreement for acceleration pulses with the peak amplitudes between 20,000 and 120,000 G (1 G = 9.81m/s(2)), rise times as short as 20 mu s, and pulse durations between 40 and 70 mu s.