Extending bubble-induced turbulence modeling applicability in CFD through incorporation of DNS understanding

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Nuclear Science and Engineering, 2018.This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.Cataloged from student-submitted PDF version of thesis.Includes bibliographical references (pages 147-153).Precise knowledge and understanding of the multiphase flow distribution is essential for light water reactor design. Multiphase Computational Fluid Dynamics (M-CFD) modeling and simulation techniques provide three-dimensional resolution of complex flow structures, which can be used to improve operation and safety in current systems, while driving optimization and performance enhancement in next generation designs. Introducing bubbles into liquid flow dramatically modifies the turbulent kinetic energy profile. Examination of experimental and Direct Numerical Simulation (DNS) research reveals a complicated, incomplete, and conflicting picture of bubble-induced turbulence (BIT). Incorporating these physical mechanisms into a BIT model compatible within the Eulerian-Eulerian framework remains a formidable challenge. Two-equation BIT models share a general formulation, manifesting as additional source terms in traditional turbulence models to account for production and dissipation of bubble-induced turbulence. Existing formulations struggle with reliably predicting the turbulent kinetic energy profile, routinely yielding non-physical results that subsequently worsen mean flow predictions. The present work encompasses two research objectives that include (1) advancing the understanding of the complex effects bubbles pose on liquid turbulence, and (2) proposing an approach to incorporate these physical phenomena into a BIT closure relation. Greater understanding of two-phase turbulent mechanisms is advanced through statistical analysis of upward bubbly channel flow DNS data generated by Bolotnov and Lu/Tryggvason. The impact of bubble deformability on the resulting turbulent distributions, energy budgets, and scales are quantified and examined. A methodology that incorporates these fundamental mechanisms into a new BIT model is proposed. The closure comprises five components that include new turbulent viscosity and time-scale formulations in addition to optimized values for the modulation parameter, dissipation coefficient, and newly proposed turbulent viscosity multiplier. Model performance and improvement is confirmed through simulation of the entire Liu (1989) experimental database and comparison with existing closures. The model is incorporated into the Bubbly And Moderate void Fraction (BAMF) formulation (Sugrue et al., 2017) in order to deliver best practices and guidelines for application of momentum closures with BIT modeling. This is accomplished through redefinition of the Wobble number, calibration of expressions for the turbulent dispersion coefficient and lift inversion function, and assessment via simulation of experimental databases.by Benjamin Lawrence Magolan.Ph. D