Output-only modal identification or operational modal analysis (OMA) has gained increasing popularity in many fields of engineering. In the context of OMA there is no need to measure the input. Modal parameters of dynamic systems are estimated just based on the output responses. One of the most robust and popular frequency domain methods of OMA is enhanced frequency domain decomposition (EFDD) method. EFDD is widely used as an interesting solution for OMA in a large number of searches, projects, and commercial software. We first assessed the EFDD features considering the design parameters selection and then introduced our recently proposed algorithm named in-operation modal appropriation (INOPMA) for use with EFDD method to improve the identification of modal frequencies and damping ratios and overcome some of the existing drawbacks. Modal identification process starts with EFDD and then INOPMA is applied on derived normalized auto-correlation functions (NACF) to estimate modal damping ratios and natural frequencies.Typical EFDD takes advantage of logarithmic decrement (LD) algorithm and zero crossing (ZC) method at this stage. A simulated fourstory shear frame has been employed to perform the evaluation of the proposed method on output-only prediction of ideal natural frequencies and modal damping ratios. The outcomes show that the modal parameters are estimated with much less bias error and variance compared to the typical EFDD procedure. The real data of the heritage court tower ambient vibration test is also used to perform sensitivity analysis of EFDD-INOPMA modal parameter estimation in the presence of measurement noise, and favorable results have been obtained.

Keywords