This paper presents a novel excitation method for in situ experimental modal analysis (EMA) and compares its performance to ambient vibration analysis, also known as operational modal analysis (OMA). The method consists of base excitation through ground accelerations generated through a large mobile shaker, a unique form of building excitation which has not been studied to date. The controllable and repeatable nature of the excitation complements some limitations of OMA, including inconsistent excitation conditions and low energy at high frequencies. The method is applied to an instrumented eleven-story building in Austin, Texas, resulting in observations of significant energy input up to 50 Hz. The high input bandwidth is essential to characterizing high order modes which ambient sources do not excite. Using a five-minute set of multiple 20-s chirp excitations from 5 to 50 Hz in two directions (vertical and transverse), it is shown that the in situ seismic EMA yields modes which are not identified through standard ambient OMA. The differences in identification methods were most noticeable above 7 Hz, where vertical contributions accounted for a significant portion of the modal response. Above 11 Hz, ambient excitation was too weak for proper OMA, but EMA enabled the identification of high-order “local” modes, although their complete identification was limited by low sensor density. The primary benefit of OMA comes from processing time series of much longer duration (20 min), which reduced variation in most of the modal estimates, given the fairly consistent ambient excitation conditions during testing. Overall, in situ seismic testing is shown to be a viable option for forced vibration testing of very large civil structures. With refinements in the testing methods, the technique could be used as a complement to ambient vibration testing in contexts where the ambient excitation bandwidth is limited or quick, short-term tests are desired.
Read full abstract