multiphase interactions which when why how
play

Multiphase Interactions: Which, When, Why, How? Ravindra Aglave, - PowerPoint PPT Presentation

Multiphase Interactions: Which, When, Why, How? Ravindra Aglave, Ph.D Director, Chemical Process Industry Outline Classification of Multiphase Flows Examples: Free Surface Flow using Volume of Fluid Examples: Eulerian Multiphase Choice


  1. Multiphase Interactions: Which, When, Why, How? Ravindra Aglave, Ph.D Director, Chemical Process Industry

  2. Outline Classification of Multiphase Flows Examples: Free Surface Flow using Volume of Fluid Examples: Eulerian Multiphase • Choice & Importance of Phase Interactions • Mesh Size Influence • Mesh Type Influence Examples: Lagrangian Models • Mesh and Turbulence Future advancements / Other models

  3. Multiphase Interactions L-L Extractors, Suspended solids, erosion Hydro-cyclones Liquid Blast furnace Separators L-L • Stirred vessel, • Liquid Bubble column (EMP) L-L-G • Offshore & Marine (VOF) S-L G-L S-S G-L-S • Coating (VOF) • Solid Icing, SCR (fluid film) G-S • Windshield (DMP) Gas Solids Cyclones, Stirred vessel, Fluidized bed Bubble Column, Pipeline flows

  4. Mixing of rubber in Banbury mixer No Slip Full Slip Partial Slip

  5. Coating d = 2.7 mm v= 4.551 m/s. Surface: waxed Contact angle advancing = 105° Contact angle receding = 95° σ = 0.073 N/m We = ρu2D/σ = 263 ( convective/surface) At wall: 6 µm Time step: 0.2 µs S. Sikalo and E. Ganic , Phenomena of droplet-surface interactions, Experimental Thermal and Fluid Science, 2006

  6. Gas – Liquid Dispersed Flow in Stirred Vessel: Geometric Setup Property Value Rushton impellers 4 Blades per impeller 6 Blade height 0.14m Blade length 0.17m Bottom clearance C b 1.12m Impeller distance C i 1.45m Impeller diameter 0.7m Liquid level H 6.55m 22m 3 Liquid volume Tank diameter T 2.09m Baffles 4 Vrabel, P. et al. (2000), Chem. Eng. Sci. 55

  7. Influence of Phase Interaction Drag! (D) Buoy oyancy ancy Buoyancy! (B) LF LF Turbulent Dispersion! WLF Drag ag VM VM Lift (LF)? Wall Lubrication (WLF)? u f Virtual Mass (VM)?

  8. Drag Force Models Linearized Standard Composed together from a Standard Drag Coefficient for a single particle Integrated models for the interphase drag or friction force. Drag Correction factor to account for multi-particle and other effects Specifying complete models for particle drag that already include Describes how concentration modifies multi-particle effects the single-particle Drag Coefficient model in a multi-particle system

  9. Overview of the Drag Force Models Linearized Standard • • Constant Constant • • Field Function Field Function • • Gidaspow (P) Schiller-Naumann (S) • • Syamlal O’Brien (P) Hamard and Rybczynski (S) • • Symmetric Drag Tomiyama (B) • Coefficient (M) Bozzano-Dente (B) • Wang Curve Fit (A) The options are qualified by the main application areas: (A) air bubbles in water systems only. (B) bubbles (M) fluid-fluid mixtures in separation applications. (P) solid particles at high concentration. (S) spherical particles at moderate concentration - including small droplets or bubbles

  10. Drag Correction Methods Bubble Regime Air / Water Non-dimensional Size Bubble Behaviors Suggested Drag Bubble Size (d) Correction Method Small spherical < 2.75 mm Eo < 1 Hindering Richardson Zaki Small ellipsoidal ~ 5 mm Eo ~ 3.3 Hindering Lockett Kirkpatrick Deforming Intermediate size Simonnet ~ 7-10 mm Eo ~ 6.6-13.4 Hindering: 0-15% void fraction Swarming: 15-30% void fraction Large spherical-cap ~ 11-14 mm We(drift velocity) ~ 8 Breakup Volume Fraction in churn-turbulent Coalescence Exponent flow Swarming

  11. Flow Pattern – Water & Gas Holdup No Aeration Aerated

  12. Mesh Independency (Polyhedral Mesh) Results are almost mesh independent even with coarsest mesh (243k cells)

  13. Influence of Bubble Size Monodisperse bubble size (1, 2 and 3mm) 450k polyhedral cells S-gamma model incl. coalescence &breakup (log.-normal distribution: 1e-4mm < BS < 10mm)

  14. Influence of Cell Type on Simulation Time 300 2000 250 t / iteration [s] Total CPU Time [h] 1500 200 150 1000 100 500 50 0 0 Hex Tet Poly Hex Tet Hex Tet Poly Hex Tet 600k 650k 453k 1.3M 2.0M 600k 650k 453k 1.3M 2.0M BUT Polyhedral cells need more time Convergence is much faster per iteration Vir Virtual ual mass ass, lift lift forc rce e & w & wall ll lu lubricat cation ion forc rce e of n neglig gligible ible im importanc ortance e in in stirr irred ed vessel ssel sim imula ulati tions ons

  15. Bubble Column Drage Force: Tomiyama Lift Force: Tomiyama Turb. Disp. Force Bubble Induced Turbulence (Troshko&Hassan) Virtual Mass Force Diaz et al. (2008), Chem. Eng. J. 139, 363-379 Ziegenhein (2013), CIT, accepted manuscript

  16. Air Buffer or Degassing? With Large Scale Interface Capturing

  17. Liquid-Liquid: Water Oil Separation Water er-Oil Oil: m in = 1.02 kg/s flow-split Acting flow-forces (0.1) 1% VF oil – Pressure-gradient – Drag & lift, – Added & virtual mass – Turbulent dispersion – Gravity Algebraic Reynolds stress model Linear/quadratic eddy-viscosity models LES/DES filtering flow-split 1.5 m (0.9) 14M trimmed cells

  18. Eulerian – Eulerian Flow Field Fully-coupled transient Eulerian-Eulerian calculations for different droplet-sizes (D) oil-water pressure journey oil volume-fraction D = 40 μm 60 μm 80 μm 100 μm oil -1.5 0 vf 0.05 water water 0 p (bar)

  19. Lagrangian Approach One-way steady-state Eulerian-Lagrangian calculations for different droplet- sizes (D) droplets distribution oil-volume fraction 0 vf 0.05 -1 z-vel (m/s) 1 D=40 0 μm 60 μm 80 μm 100 μm

  20. Validation Efficiency ( η ) Droplet diameter (µm) η =100*(1-m out /m in ) m out : is the oil mass exiting from the clean outlet (top) m in : is the total oil mass imported in the hydrocyclone

  21. Eulerian Multiphase Large Scale Interface (LSI) Model Elimnates the need of VOF with extremely fine mesh to resolve bubbles and droplets Captures many different co- existing flow regimes – Wave formation due to interfacial shear – Spray generation – Droplet carry-over by the gas flow – Bubble entrainment into liquid – Slug flow – Stratified flow / free surfaces – Dispersed sprays – Dispersed bubbles Gas-Liqui iquid d Counter ter-Cur urrent rent flow in PWR [Deendar ndarli liant anto et al., NED, 39 (2012)] )]

  22. LSI: A simple flow topology detection method   d threshold i threshold , , Variation of Blending Function with Volume Fraction of reference Dispersed Phase 1 Blending Function Bubbly Dispersion 0.5 Inversion Droplet Free Surface Transition 0 (free surface) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Volume Fraction of reference Dispersed Phase

  23. LSI – Interface Turbulence Damping Air-Water Stratified flow experiment of Fabre et al. (1987)  12m long horizontal channel. Approx. Re=40,000.  Cross-section : 10 cm high x 20 cm deep.  2D grid (12m x 0.1m) : 400 x 54 Pressure Drop (Pa/m) Experimental 2.1 LSI – No interface 19.68 damping LSI – with interface 2.63 damping Air (3.66 m/s) For r Inter ernal nal Use e Only ly Water (0.395 m/s)

  24. LMP-VOF LMP->VOF Impingement, new feature in STAR-CCM+ v10.02 VOF->LMP Stripping, currently under development, targeting STAR-CCM+ v10.04/10.06

  25. VOF - Fluid Film Interaction Model D881 Locally chooses the most suitable model for the local flow regime Jet (VOF OF) Thin Film (Fluid uid Film) m) Thick Film (VOF)

  26. Trickle Bed Reactors Trickle Bed reactors – VOF-Fluid Film Interaction – Packed bed modeling approach Edge stripping with Wave stripping with Fluid film Multiple VOF film formation fluid film fluid film particles

  27. Conclusions Breadth & Flexibility • Breadth + Flexibility + Best Practices = SUCCESS! Expanding Phase Solve wide Mesh Size Mesh Type Degassing vs. model Interaction range of Influences Influences Air Buffer compatibilitie Parameters problems s

Recommend


More recommend