Shock and Vibration Application Brief

Bridge Vibration

The Solution

The research team uses DADiSP, the graphic display and data processing software by DSP Development Corporation, to support their work on bridge monitoring.

Bridge Signature

The University of Connecticut model bridge studies have demonstrated that a bridge has a "signature," as individual as a human fingerprint, which includes its natural frequencies and mode shapes. (A mode shape is a discrete shape assumed by a bridge at a natural frequency of vibration.) Changes in the signature correspond to changes in the bridge's structural stiffness. The mass, velocity, and position of a vehicle crossing a bridge do not have a significant influence on its natural frequencies and mode shapes; therefore, monitoring these variables over time to evaluate structural condition has so far seemed promising.

Natural Frequencies

The monitoring system by Vibra-Metrics consists of sixteen accelerometers, two cluster boxes, and a sentry unit which houses a computer. The accelerometers - sensors that detect vibrations - are magnetically attached to the bridge's girders and positioned throughout. To establish the bridge's signature, time history plots are obtained from the vibrating bridge and transformed into the frequency domain through software at the monitoring system. The frequency spectra are imported into DADiSP, where they are analyzed to determine which natural frequency peaks can be used for monitoring.

Frequency Peaks

A new DADiSP labbook is created for each data acquisition time period. Three worksheets are created with six windows each. Each worksheet represents a group of accelerometers located either along the bridge's girders or across its centerspan. DADiSP's ZOOM and OVERPLOT commands help identify potential natural frequency peaks common to all accelerometers. These common peaks correspond either to natural frequencies or to system noise; they can be differentiated because a natural frequency peak has an associated mode shape and a system noise peak does not. Phase studies using DADiSP's cross-correlation macro FCROSS are performed to define a mode shape, then spectrum clean-up techniques are used to enhance the appearance of the frequency spectrum. Windowing functions, for example, help remove FFT leakage, which tends to widen natural frequency peaks, causing misinterpretation.

DADiSP Easily Processes Large Data Sets

Dean Bagdasarian likes DADiSP's ability to handle large data sets and its capacity for storage. He feels that using it makes his work easier. He states, "DADiSP provides an easy way to view and analyze the tons and tons of data we have coming in from hundreds of different data collection points at a fast pace. It helps us analyze and separate them and keep everything organized." He says he also appreciates DSP Development's technical support staff, who helped his research team overcome compatibility problems. Overall, he has found DADiSP to be a valuable tool "in every function - we use so many that I don't have a favorite." Ultimately, he and his team, with DADiSP's help, will find a way to ensure bridge safety so that we won't need collapses to show us which bridges need fixing.