Quantitative Validation of Gas-Liquid Flow Regime Transition Using Eulerian-Eulerian CFD Models

Abstract

Quantitative validation of a computational fluid dynamics (CFD) code is about making point-to-point comparisons over fields of data to make statements about predictive fidelity. By the very nature of CFD, these comparisons are high-performance computing and data intensive. This presentation provides an overview of workflow development toward quantitative development of two-phase CFD codes. The focus here is on multifield two-fluid model capability implemented in the commercial CFD code ANSYS CFX applicable to boiling gas-liquid two-phase flows. Applications to validating three-dimensional predictions that span two-phase flow regimes are discussed.

Keywords:

• Two-fluid model
• computational fluid dynamics
• gas-liquid two-phase flow

Acknowledgments

In addition to the coauthors, I am thankful for other peers and university and commercial partners. At the Naval Nuclear Laboratory, Steve Beus, Frank Lincoln, Richard Bauer, and Hyung Lee were formative in my understanding of the two-fluid model and practical application of CFD codes. Kevin Hogan, Dart Strayer, and Gretchen Burdick have been trusted colleagues in putting CFX through its paces.

Professor Mamoru Ishii of Purdue University is foundational to this work from concept to execution and I am indebted to his efforts. Professor Martin Lopez de Bertodano of Purdue University has been a willing partner expanding what is possible with the two-fluid model and CFD. University collaboration partners have been very good to me through the years and with a long enough time integral I have worked with many, including Seungjin Kim and Takashi Hibiki at Purdue University, Yang Liu at Virginia Tech, Xiaodong Sun and Steve Ceccio at University of Michigan, Simo Mäkiharju at University of California Berkeley, Caleb Brooks at University of Illinois, Josh Schlegel at MUST, Subash Sharma at University of Massachusetts Lowell, and Barton Smith at Utah State.

Barbara McFarland of McFarland Technologies has been instrumental in my understanding of data relationships and facilitating fruitful discussions with Bill Oberkampf and John Mahaffy.

None of this would be possible without the patience and expertise through the years of Paul Guilbert, Alan Burns, Christiane Montavon, and Conxita Lifante have been essential partners at ANSYS making CFX a capable, robust tool.

Disclosure Statement

No potential conflict of interest was reported by the authors.