A novel entropy normalization scheme for characterization of highly compressible flows

Ethan A. Vogel, James G. Coder

Research output: Contribution to journalArticlepeer-review

5 Scopus citations

Abstract

Abstract: A matrix normalization scheme based on thermodynamic entropy is derived for modal decomposition techniques applied compressible flows. It is demonstrated that this normalization scheme is consistent with the scalar form of entropy. Analysis based in this consistency is performed to demonstrate the theoretical underpinnings of the Chu energy norm, which is the industry standard for compressible modal decompositions. The entropy normalization is shown mathematically to converge to the Chu normalization in the absence of strong temperature gradients. It is then compared to the Chu normalization by analyzing transient growth calculated through a linear stability analysis of a self-similar compressible boundary layer profile. The entropy norm is shown to be more sensitive to temperature fluctuations around the boundary layer than the Chu norm. These observations are further validated in a POD implementation of the entropy normalization, contracted about the conservative and primitive variables. The trends observed in transient growth analysis are observed in full-scale POD. The potential for the entropy normalization to be applied to flows with additional relevant physics, such as as thermal and chemical nonequilibrium, is explored. Graphical abstract: [Figure not available: see fulltext.].

Original languageEnglish (US)
Pages (from-to)641-670
Number of pages30
JournalTheoretical and Computational Fluid Dynamics
Volume36
Issue number4
DOIs
StatePublished - Aug 2022

All Science Journal Classification (ASJC) codes

  • Computational Mechanics
  • Condensed Matter Physics
  • General Engineering
  • Fluid Flow and Transfer Processes

Fingerprint

Dive into the research topics of 'A novel entropy normalization scheme for characterization of highly compressible flows'. Together they form a unique fingerprint.

Cite this