Large Scale Vibration Assessment Utilising a Hybrid Measurement Approach to Inform Detailed Finite Element Modelling

Abstract Flow-Induced Turbulence (FIT) is one of the most common root causes for fatigue failures in process pipework. While recent advancements have improved design-phase guidance to mitigate FIT, addressing these issues in existing plants remains challenging and can be cost-prohibitive due to the need for major changes in process conditions, design, or physical layout. A novel approach to analyse and ensure the safe operation of plant equipment and structural arrangements has been developed for environments where traditional strain measurements are impractical. This approach includes the use of advanced imaging, in combination with traditional handheld analysis tools, to describe the system behaviour and response for multiple large-scale structures. This enables results of the site assessment to be fed into an advanced numerical analysis with the aim to provide a complete integrity assessment of locations which would be not be otherwise assessable. Using a case study, it is shown how the risk of vibration-induced fatigue (VIF) damage in a complex piping system and its structural support is assessed through the application of Finite Element Models (FEMs). Models are refined using measurements from a vibration survey and data from a motion amplification camera (MAC) to produce structural deflection shapes. These are then used in a numerical analysis to create a representation of the as-built arrangement and its vibration response. The numerical analysis uses a series of detailed submodels to enhance model fidelity for areas with concerning predicted dynamic stress responses and extracts dynamic stresses at critical locations. The results are then compared to the relevant standards and guidelines. This hybrid measurement approach, combining advanced imaging techniques and FEM, provides a robust framework for assessing and ensuring the structural integrity of complex systems under dynamic loading conditions.