A Comparative study of the methods for scaling of mode shapes in operational modal testing

The objective of Operational Modal Analysis (OMA) is identification of the modal properties using only the measured responses without the knowledge of input forces. In OMA, the mode shapes cannot be scaled straightforward mainly because the excitation force is not measured. However, the mass normalized mode shapes are required in many applications such as response prediction, stress analysis or damage detection. If the structure is modified by adding extra mass or stiffness to the structure and a new test is carried out, the scaling factors can be obtained. There are several relations for estimating the scaling factors, based on the mass change or mass-stiffness changes. The accuracy of these equations depends on the mass-stiffness change magnitude, the number and the location of the mass and stiffness attached to the structure. In this article a comparison is made between different approaches for scaling the unscaled mode shapes from OMA. The simulation of a clamped-clamped beam was carried out and some suggestions are made in order to decrease the scaling error.