Application of a Pre-Scaling Assembly Approach in the Automation and Improvement of Global Mode Shapes Computation through Subspace-Based Modal Analysis

Operational modal analysis (OMA) techniques have become very popular in the characterization of dynamic properties of civil engineering structures. This is mainly because of their practical capabilities and implementation economy. For civil engineering applications, OMA aims to identify natural frequencies and the related modal damping and mode shapes of a structure under ambient vibration conditions. Generally, when identifying mode shapes, the number of measurement points is an important factor which is usually restricted by instrumentation limitations. To overcome this limitation, data acquisition in multiple patches with roving sensor setups can be applied. In this case, due to the non-stationary nature of ambient vibration, different parts of identified mode shapes are not scaled and the parts of mode shapes should be properly scaled before assembly. The primary objective of this study is to implement the pre-scaling assembly technique in reference-based stochastic subspace identification (SSI/Ref). This procedure allows the automation of the modal identification process in cases of multiple non-simultaneously measured data and takes advantage of the non-parametric assembly technique while using the SSI method. In order to estimate the effectiveness of the applied technique, a three-span steel bridge structure (I-386 Bridge located in Arlington, TN) is considered as the experimental model. Post-scaling and pre-scaling techniques are applied for the identification of mode shapes and the results are compared. Results of the study show that the applied technique can effectively estimate the modal properties of a structure and can be simply implemented and automated.