7.1.7. Large-Scale Jet Flames Modeling
Last updated
Was this helpful?
Last updated
Was this helpful?
Version: GASFLOW-MPI 2.0 Revision 4808 (October 18, 2024) or a more recent revision.
Fuel Type: We currently support hydrogen and methane as fuel types. However, new fuel types can be easily implemented to meet user requirements.
Note: The value of the input parameter, iburn, for the combustion models is currently dependent on the code development progress, which lacks a logical order. To ensure version compatibility, the existing options for iburn will remain unchanged, and any new model options will be added incrementally. As user requirements evolve, the combustion models for new fuel types will be adapted and expanded accordingly. In most cases, each value of iburn will be limited to a specific fuel type to maintain compatibility and simplicity. However, options for general fuels can also be provided to model chemical kinetics, turbulent combustion, and turbulence-chemistry interaction.
Activate the Eddy Dissipation model for hydrogen jet flame modeling
iburn = 3, ; activate the Eddy Dissipation model for hydrogen jet flame modeling
Activate the Eddy Dissipation model for methane jet flame modeling
iburn = 7, ; activate the Eddy Dissipation model for methane jet flame modeling
In the meanwhile, inonpremix has to be switched on to model the jet flames.
inonpremix = 1, ;for non-premixed combustions used when iburn = 3 or 7
An ignitor model must be also defined.
Ignition starts at 0.0 s and the duration is 0.2 s.
xi_ignitdef(1:10,1) = 5, 6, 1, 2, 23, 24, 1, 0.0, 0.2, 0, ; Ignitor Model
Ignition starts at 0.3 s and the duration is 0.05 s.
xi_ignitdef(1:10,1) = 5, 6, 1, 2, 23, 24, 1, 0.3, 0.05, 0, ; Ignitor Model
Note: When using the Eddy Dissipation model, the combustion is assumed to be infinitely fast, which means the fuel and oxidizer react instantaneously as soon as they are mixed, without any delay. This simplification can lead to inaccuracies in situations where the combustion process is controlled by the chemical kinetics, rather than the turbulent mixing. For example, if the ignition is delayed by 0.3 seconds in the example bove, the Eddy Dissipation model would predict the mixtures to be burned immediately as soon as the ignition starts. However, this may not be an accurate representation of the actual combustion behavior, especially for cases with significant ignition delay or slow chemical reactions. For example, it may overpredict the pressure and thermal radiation due to the very fast combustion.
To address this limitation, a better approach is to combine the finite-rate chemical kinetics model with the Eddy Dissipation model. This hybrid approach can more accurately capture the combustion process when the chemical kinetics dominate the overall behavior, such as in transient simulations.
On the other hand, for steady-state simulations of jet flames, the combustion is often dominated by the turbulent mixing rather than the chemical kinetics. In such cases, the Eddy Dissipation model alone may be sufficient, as the infinite-rate combustion assumption would be a reasonable approximation.
Gas spieces for hydrogen jet flame
mat = 'h2', 'o2', 'n2', 'h2o', ; gas spieces for hydrogen jet flame
Gas spieces for methane jet flame
mat = 'ch4', 'o2', 'n2', 'h2o', 'co2', ; gas spieces for methane jet flame
Turbulence model as well as mass/momentum/energy diffusion must be switched on in jet flame simulations.
It is crucial to set up appropriate initial and boundary conditions for the turbulence model. If the turbulence conditions are not configured with reasonable values, the time step may become very small, which can significantly impact the simulation performance.
In the example below, the initial turbulence conditions (0 s) in the entire computational domain is set up in turbdef(1:12,1), and the turbulent boundary conditions (0-9999 s) at the jet nozzle is given by turbdef(1:12,2).
It is recommended to limit the CFL (Courant-Friedrichs-Lewy) numbers for advection (cflnum) and diffusion (cfldiff) to around 0.25 to ensure numerical stability. The user can set these values between 0 and 1. Using a bigger CFL number will result in a larger time step, which can reduce the overall simulation time. However, this approach must be applied with caution, as excessively high CFL numbers can also lead to numerical instability in the simulation.
The user should increase the CFL values gradually and monitor the simulation closely for any signs of instability, such as diverging results or unphysical oscillations. It is important to strike a balance between computational efficiency and numerical stability to obtain accurate and reliable simulation results.
This example demonstrates how to set up GASFLOW-MPI to model vertical and horizontal CH4-air jet flames. The user will need to refine the mesh and set appropriate initial and boundary conditions for their specific application.
