Experimental Modal Analysis

The goal of an Experimental Modal Analysis is determination of a structure’s dynamic characteristics. These characteristics are described by the structure’s natural modes. A natural mode is described by its frequency, mode shape and damping coefficient (Fig. 1).

Fig. 1: Langensand Bridge Luzern: Mode No. 4, Frequency f = 3.52 Hz, Damping zeta = 0.88%.

To perform an Experimental Modal Analysis two methods are available today:

Forced Vibration Testing, FVT, and Ambient Vibration Testing, AVT.

With FVT , the structure’s transfer function H is determined through measuring the excitation x(t) and the response y(t) (Fig. 2a).
With AVT, responses y(t) are measured for this purpose only. References R have to be kept constant throughout the test procedure (Fig. 2b).

FVT Balken Grafik AVT Balken Grafik
Fig. 2a: Forced Vibration Testing, FVT Fig 2b: Ambient Vibration Testing, AVT

Based on a large number of experimental investigations on civil engineering structures, Reto Cantieni presented an overview of the developments of these technologies since 1985 at IOMAC in 2005 (IOMAC paper, IOMAC ppt-presentation).

Until 1993, the FVT method having been developed by mechanical engineers (after the FFT algorithm having been published by Cooley/Tukey in 1965) existed only. At that time, among others, EMPA also applied this method to civil engineering structures. This, however, required construction of quite large servo-hydraulic shakers requiring substantial amounts of pressurized oil and electric energy.

AVT has been developed by civil engineers. There were three reasons for this: a), it is not always easy to dispose of an artificial source of excitation for large, low-frequency structures, b), it is not always possible to close a structure to normal operation for a test, and, c), application of artificial shakers is usually quite expensive.

In contrary to FVT, the modal parameters derived from AVT are unscaled, i.e. no information on the structure’s mass and stiffness is included. This means among others that automatic computer-based updating of FE models is not possible without work-arounds.

In 1993, Andreas Felber published his PhD thesis. It became then possible to apply AVT procedures to large structures including many measurement points with a reasonable effort. During a research stay at EMPA in 1994/95, Felber successfully applied his software in seven AVT-tests on civil engineering structures of different size in Switzerland. The respective report is available at rci dynamics. However, subsequent commercial application of the Felber software kept in quite narrow limits.

In 2000, the today widely used ARTeMIS software package came on the market (www.svibs.com; Artemis Flyer). Since the beginning, Reto Cantieni has been a Technical Advisor for this software package. rci dynamics is also an ARTeMIS distributor since 2006.

Reto Cantieni disposes of one of the world-wide largest experiences with practical application of ARTeMIS to civil engineering structures (Wind Turbine Gossau, Dancing Locarno, Eglisau Power Station, Erfttal ICE-Railway Bridge, San Luzi Bell TowerLangensand Highway Bridge). (Publications, Downloads)