Modal analysis of high-fidelity simulations in turbomachinery

Christian Morsbach,Bjoern F. Klose,Michael Bergmann,Felix M. Möller
2024-03-01
Abstract:We revisit recently published high-fidelity implicit large eddy simulation datasets obtained with a high-order discontinuous Galerkin spectral element method and analyse them using Proper Orthogonal Decomposition (POD) as well as Spectral Proper Orthogonal Decomposition (SPOD). The first configuration is the MTU T161 low-pressure turbine cascade with resolved end wall boundary layers in a clean version and one with incoming turbulent wakes. We focus on the behaviour of the laminar separation bubble and the secondary flow system and how these phenomena are affected by incoming wakes. The second configuration is a transonic compressor cascade in which we analyse the unsteady behaviour of the shock wave boundary layer interaction. Throughout the paper, we try to discuss not only the flow physics but also insights into how the modal decomposition techniques can help facilitate understanding and where their limitations are.
Fluid Dynamics
What problem does this paper attempt to address?
This paper mainly discusses how to analyze the flow phenomena in the turbine machinery of aviation engines using high-fidelity large eddy simulations (LES) data. The research team used the Discontinuous Galerkin Spectral Element Method (DGSEM) to generate data and employed Proper Orthogonal Decomposition (POD) and Spectral Proper Orthogonal Decomposition (SPOD) techniques to analyze these data. The paper first points out that with the increase in computational resources and the development of advanced methods, high-fidelity fluid dynamics data has become more complex, making the analysis of these data challenging. The researchers used POD and SPOD to extract the most energetic modes and sorted them by frequency content to understand the flow physics, while also discussing the limitations of these methods. Specifically, the paper analyzes two configurations: the flow of the MTU T161 low-pressure turbine blade, focusing on the influence of turbulent wakes on boundary layer separation bubbles and secondary flow systems; and the shock boundary layer interaction in a transonic compressor stage. Through SPOD, the study found that wakes amplify certain modes and increase the turbulence at downstream edges, while in the compressor stage, the instability of the shock wave-boudary layer interaction was analyzed. In addition, the paper compares the advantages and disadvantages of POD and SPOD in handling data, pointing out the superior frequency resolution of SPOD, which can distinguish phenomena at different frequencies more clearly and avoid misleading conclusions that may be caused by the lack of frequency resolution of POD. Through these methods, researchers are able to gain a deeper understanding of complex flow phenomena. Overall, this paper aims to address how to effectively analyze high-fidelity fluid dynamics data in order to better understand and predict the flow behavior in turbine machinery.