LES of Spray Combustion using Flamelet Generated Manifolds

Transcription

1 LES of Spray Combustion using Flamelet Generated Manifolds Armin Wehrfritz, Ville Vuorinen, Ossi Kaario and Martti Larmi Aalto University Thermodynamics and Combustion technology April 07, 2014

2 Part I: Non-reacting spray study Objectives 1. Mesh resolution effects in Large Eddy Simulation 2. Influence of droplet breakup modeling on the local and global flow characteristics Spray case Non-reacting Spray A baseline: P inj = 150 MPa T = 900 K P amb = 6 MPa 0 % O 2 Reference, V. Vuorinen, O. Kaario, and M. Larmi., Large Eddy Simulation of High-Velocity Fuel Sprays: Studying Mesh Resolution and Breakup Model Effects for Spray A. Atomization and Sprays, 23(5): , LES of Spray Combustion using FGM 2/15

3 Computational Methods Gas phase Large Eddy Simulation (LES) Turbulence modeling based on the implicit LES approach [2] No explicit subgrid scale model Finite Volume, open source CFD code OpenFOAM 2.0.x (2nd order accurate in space and time) Liquid phase Lagrangian Particle Tracking (LPT) No explicit primary break-up model Initial droplet size distribution (Rosin-Rammler) Secondary break-up models: 1. Enhance Taylor Analogy Breakup (ETAB) 2. Kelvin-Helmholtz Rayleigh-Taylor (KHRT) LES of Spray Combustion using FGM 3/15

4 Computational mesh Fully hexahedral Refinement in the spray region by 2:1 cell splitting Applied cell sizes: dx [µm] N cells [-] M M M M Constant time step: t = s LES of Spray Combustion using FGM 4/15

5 Results Liquid length Poor results for 250 µm cell size meshes, regardless of breakup model 125 µm cell size: ETAB: Good agreement with experiments KHRT: Significantly over-predicted Good results for 62.5 and µm cell size meshes, regardless of breakup model (a) ETAB (b) KHRT Liquid penetration [mm] Liquid penetration [mm] Experiments 250 µm µm 62.5 µm µm Time [ms] Experiments 250 µm µm 62.5 µm µm Time [ms]

6 Results Mixture fraction distribution (KHRT model) 250 µm cell size mesh not able to capture the turbulent motion correctly 125 µm cell size mesh able to capture a significant part of the turbulent motion Increasing level of detail for 62.5 and µm cell size meshes 250 µm 125 µm 62.5 µm µm

7 Results Vapor penetration (KHRT:, ETAB: no marker) Slight under-prediction for all cell sizes and both breakup models Vapor penetration [mm] Experiments 125 µm 62.5 µm µm Radial mixture fraction profile (KHRT:, ETAB: no marker; z = 25 mm) 125 and 62.5 µm mesh: Values in the the center is under- and spreading over-predict µm mesh: Good agreement with experiments Vapor mass frac. [-] Time [ms] Experiments 125 µm 62.5 µm µm r [mm]

8 Results PDF of droplet diameter ETAB 125 µm 62.5 µm µm ETAB model Uniform distribution SMD = 0.3 µm µm PDF KHRT KHRT model Broad range of droplet sizes SMD = 1.1 µm µm d [µm] Probability density function of droplet diameter at t = 1.4 ms

9 Part II: Non-reacting spray study Objectives: Investigate the ignition characteristics and and early flame structure using Large Eddy Simulation and Flamelet Generated Manifold (FGM) Spray case Reacting Spray A cases: P inj = 150 MPa T = 900 K P amb 6 MPa ρ amb = 22.8 kg/m 3 15 % O 2 LES of Spray Combustion using FGM 9/15

10 Computational Methods Flow solver Implicit Large Eddy Simulation Lagrangian Particle Tracking (Secondary breakup: ETAB) OpenFOAM 2.2.x Advanced thermodynamic/transport models (i.e. Wilke/Mathur mixture models) Flamelet Generated Manifolds (FGM) [3] Tabulated chemistry model State of combustion is parametrized by a few control variables (here, mixture fraction and a reaction progress variable) Chemistry data obtained from 1D igniting/steady counterflow diffusion flames (i.e. flamelets) Detailed chemical kinetics (253 species, 1437 reactions [4]) LES of Spray Combustion using FGM 10/15

11 FGM tables C [-] Z [-] C [-] Z [-] Figure : Temperature Figure : CO mass fraction Chemistry parametrized by mixture fraction Z and reaction progress variable C LES of Spray Combustion using FGM 11/15

12 Results Spray penetration Calculation are carried out on the 62.5 µm mesh Simulated liquid length matches the experimental data Vapor penetration slightly under-predicted Spray penetration [mm] Experiments Simulation Time [ms]

13 Results Ignition delay Significant over-prediction compared to experiments Consistent ignition delay estimate for both ECN definitions ( dt dt & OH mass fraction) Flame length Lift-off length slightly under-predicted T [K] Spray flame penetration [mm] Temperature t [ms] Experiments Lift-off length Jet penetration (Zst = 0.045) Ignition OH Experiments 1e t [ms] 8 6 OH [-] 4 2 0

14 Results Spray A combustion

15 Questions & Discussion Thank you for your attention! LES of Spray Combustion using FGM 15/15

16 1 Appendix LES of Spray Combustion using FGM 1/4

17 1 References [1] Armin Wehrfritz, Ville Vuorinen, Ossi Kaario, and Martti Larmi. Large Eddy Simulation of High-Velocity Fuel Sprays: Studying Mesh Resolution and Breakup Model Effects for Spray A. Atomization and Sprays, 23(5): , ISSN doi: /AtomizSpr URL 6a7c7e cc,67b312a93f7dc969, 0a7117ff52a01272.html. [2] F. F. Grinstein, L. G. Margolin, and W. J. Rider. Implicit Large Eddy Simulation. Cambridge University Press, ISBN LES of Spray Combustion using FGM 2/4

18 1 References II [3] J. A. van Oijen and L. P. H. de Goey. Modelling of Premixed Laminar Flames using Flamelet-Generated Manifolds. Combustion Science and Technology, 161(1): , December ISSN , X. doi: / URL tandfonline.com/doi/abs/ / LES of Spray Combustion using FGM 3/4

19 1 References III [4] Krithika Narayanaswamy, Perrine Pepiot, and Heinz Pitsch. A chemical mechanism for low to high temperature oxidation of n-dodecane as a component of transportation fuel surrogates. Combustion and Flame, 161(4): , April ISSN doi: /j.combustflame URL S LES of Spray Combustion using FGM 4/4