Special Transient Testing Windows. In the early ’70s, the most
important signal processing development for impact testing was
the development of force and exponential windows. This development was the fallout from work being done by Ron Potter, who
was the person at HP primarily responsible for the development of
the HP 5450 series Fourier analyzer system. He was the real guru
of Fourier analysis in the late ’60s and ’70s. As part of one of his
activities, he was developing a parameter estimation algorithm for
extracting modal parameters. He was trying to get starting values
for the modal parameters, using the shift theorem of the Fourier
transform to reduce the apparent damping in his FRF measure-
ments. The shift theorem states that multiplying a time function
by an exponential function will shift the damping or frequency
axis in the transformed domain. In other words, the apparent
damping of a system can be changed in a predictable way simply
by multiplying the unit impulse response measurement of a system
by a damped exponential. It became apparent that this could be
used as a window to eliminate “leakage errors” and to improve
the signal to noise ratio of impact measurements. As result, an
exponential window could be applied to both the input force and
the output response signal. An additional force window could be
applied to the force signal to eliminate noise on the force channel
after the impact. These windows have been documented in a
number of references.
4,5
Special Transient Testing Windows. In the early ’70s, the most important signal processing development for impact testing was the development of force and exponential windows. This development was the fallout from work being done by Ron Potter, who was the person at HP primarily responsible for the development of the HP 5450 series Fourier analyzer system. He was the real guru of Fourier analysis in the late ’60s and ’70s. As part of one of his activities, he was developing a parameter estimation algorithm for extracting modal parameters. He was trying to get starting valuesfor the modal parameters, using the shift theorem of the Fourier transform to reduce the apparent damping in his FRF measure-ments. The shift theorem states that multiplying a time function by an exponential function will shift the damping or frequency axis in the transformed domain. In other words, the apparent damping of a system can be changed in a predictable way simply by multiplying the unit impulse response measurement of a system by a damped exponential. It became apparent that this could be used as a window to eliminate “leakage errors” and to improve the signal to noise ratio of impact measurements. As result, an exponential window could be applied to both the input force and the output response signal. An additional force window could be applied to the force signal to eliminate noise on the force channel after the impact. These windows have been documented in a number of references.4,5
การแปล กรุณารอสักครู่..