Location

London

Event Website

http://www.csce2016.ca/

Description

The Random Decrement Technique (RDT) has been widely used to extract as-built structural dynamic properties of civil engineering structures under ambient excitation, such as natural frequency, damping ratio, and their nonlinearity. This paper aims to clarify firstly that the RDT itself is not a damping evaluation technique (DET), but rather a data conditioning technique, akin to a filter. It results in what is called a “random decrement signature” (RDS) that is considered to represent the free decay response of the system, mode, or DOF being investigated. This paper also aims to show that a number of parameters influence the outcome of the RDT and the chosen DET. This was done by generating different sets of synthetic data, for which the actual damping and frequency values are known, which in turn are analyzed using RDT and appropriate DETs, and the results are then presented and discussed. Three important findings are: that damping may typically be overestimated especially the higher the noise is, any type of data filtering greatly affects the results, and some amplitude dependency may appear even if there was none. A discussion on how these potential pitfalls in the practical application of the RDT is then offered.


Share

COinS
 
Jun 1st, 12:00 AM Jun 4th, 12:00 AM

STR-896: POTENTIAL PITFALLS IN THE PRACTICAL APPLICATION OF THE RANDOM DECREMENT TECHNIQUE

London

The Random Decrement Technique (RDT) has been widely used to extract as-built structural dynamic properties of civil engineering structures under ambient excitation, such as natural frequency, damping ratio, and their nonlinearity. This paper aims to clarify firstly that the RDT itself is not a damping evaluation technique (DET), but rather a data conditioning technique, akin to a filter. It results in what is called a “random decrement signature” (RDS) that is considered to represent the free decay response of the system, mode, or DOF being investigated. This paper also aims to show that a number of parameters influence the outcome of the RDT and the chosen DET. This was done by generating different sets of synthetic data, for which the actual damping and frequency values are known, which in turn are analyzed using RDT and appropriate DETs, and the results are then presented and discussed. Three important findings are: that damping may typically be overestimated especially the higher the noise is, any type of data filtering greatly affects the results, and some amplitude dependency may appear even if there was none. A discussion on how these potential pitfalls in the practical application of the RDT is then offered.

http://ir.lib.uwo.ca/csce2016/London/Structural/56