On the use of operational transmissibilities for the predication & analysis of coupled assemblies

McGee, Ramin

Abstract

As is usually the case in many scientific disciplines, prediction and analysis techniques are firstly developed within laboratory conditions and subsequently extended to industrial settings through further studies. Within vibro-acoustics, we usually begin testing a method in a controlled environment via Experimental Modal Analysis (EMA). One example of an EMA-based method called the Round-trip (RT) method has gained traction within industry recently, a highly convenient and accurate driving-point mobility prediction technique which may be used in influence structural design. Often within various industry, the driving-point mobility may be needed at locations such as mounting points between two sub-structures coupled to one another. It is common that these areas are access limited, meaning conducting a measurement using a modal hammer can be impractical. The RT provides a solution to this by allowing the user to measure three transfer mobilities which uses force inputs on areas of the coupled structure that are much easier to apply.

However, a limitation facing EMA-based methods is that it requires any sources within the coupled system to be shut down. Additionally, if the force and measured response locations are far away from each other, this will yield data with poor signal-to-noise (SNR). This often limits the RT to smaller scale structures which can be tested in laboratory conditions. There has been a lack of investigation into incorporating an Operational Modal Analysis (OMA) approach into the RT method to circumvent these limitations. OMA similarly allows the extraction of modal properties, but does so by using ambient, unmeasured, and stochastic excitations (such as wind) and output-only responses.

In this thesis, we show that output-only/operational transmissibilities can be used to represent two of the three transfer mobilities in the point RT identity, allowing its extension to much larger structures and uncontrolled active environments. This novel approach to the RT, termed the Operational Round-trip (ORT) method, is analysed and compared to the original technique across three experimental examples. It is shown that it too can accurately predict driving-point mobilities while adding further convenience. Additionally, this thesis has investigated an OMA-based approach for analysing transmission paths using output-only transmissibilities. It is known that by using the ‘bottleneck’ effect and applying the SVD to a transfer mobility of a coupled system, the singular values can be analysed to detect unaccounted flanking and the number of transmission paths. Similarly to the RT method, this EMA-based method may present challenges on larger scale structures, and/or if an active component (such as a compressor) cannot be shutdown. Instead, it is shown that by identically analysing the singular values of output-only transmissibilities instead, this method can be extended to industrial applications.

The findings in this thesis show that using output-only transmissibilities provides a more convenient means of using the vibro-acoustic prediction and analysis methods presented, while also allowing them to be used in a wider range of systems within industry. Finally, an output-only extension to these methods allows the potential to be developed into real-time monitoring tools.

• Thesis Type: Thesis
• Online Publication Date: May 29, 2025
• Deposit Date: May 1, 2025
• Publicly Available Date: Jun 30, 2025
• Award Date: May 29, 2025