Improving the accuracy of two-fluid sub-grid modeling of dense gas-solid fluidized flows

• Published in: Chemical engineering science 2021-01, Vol.229, p.116021, Article 116021
• Main Authors: Niaki, Seyed R.A., Mouallem, Joseph, Chavez-Cussy, Norman, Milioli, Christian C., Milioli, Fernando E.
• Format: Article
• Language: English
• Subjects: Dense flows; Filtered closures; Fluidization; Sub-grid models; Two-fluid model
• ISSN: 0009-2509; 1873-4405
• DOI: 10.1016/j.ces.2020.116021

Drag coefficient correction and residual pressure of gas and solid phases as a function of filtered solid volume fraction and filtered kinetic energy of solid velocity fluctuations (with and without interparticle friction), for particular values of filtered slip velocity, macro-scale gas Reynolds number and solid fraction, and filter size. [Display omitted] •Large scale two-fluid simulations of gas-solid fluidized flows require sub-grid closure formulation.•Looking for that, highly resolved simulations were practiced under microscopic two-fluid modeling.•A range of macro-scale conditions and a three-independent variable correlation approach were enforced.•Accounting for macro-scale conditions and for three independent variables considerably added to accuracy. In gas-solid fluidization highly resolved simulations (HRS) with microscopic two-fluid modeling (mTFM) are frequently employed for deriving sub-grid closures for filtered parameters that are required in filtered formulations useful in large scale simulations of real flows. This work intends to contribute for improving the accuracy of underlying HRS raw data upon which the accuracy of sub-grid closures relies on, particularly regarding dense gas-solid fluidized flows. The accuracy of mTFM is improved by incorporating interparticle frictional effects. The accuracy of correlation to filtered parameters is improved by accounting for macro-scale conditions and by enforcing a three-independent variable approach. HRS were performed in a periodic domain applying the mTFM of MFIX properly modified by the inclusion of a literature trustworthy interparticle friction model. Results showed that accounting for macro-scale conditions and for three independent variables considerably contributes for improving correlation resolution, while the effect of interparticle friction was only marginal.