Waveform analysis techniques for interpreting wind and bridge motion data

Abstract

To accurately measure and properly analyze random time histories, it is necessary to give careful consideration to many technical issues. For the measurement task, sensor performance must be properly calibrated and maintained, sampling rates must be high enough and sampling periods long enough, and inputs must be correctly filtered. For analysis, the computer software must accommodate and correct for imperfect sensor performance and be able to demean, detrend, filter, band, taper, window, transform, and integrate random time histories from many inputs. In this paper the key aspects of waveform analysis techniques employed by FHWA will be presented.