燃烧;数值模拟
Reacting Flow Tips and Tricks
Advanced Combustion Modeling Course
© 2009 ANSYS, Inc. All rights reserved.
ANSYS, Inc. Proprietary
燃烧;数值模拟
Advanced FLUENT Training Combustion June 2009
General GuidelinesStart in 2DDetermine applicability of model physics Mesh resolution requirements (resolve shear layers) Solution parameters and convergence settings
Boundary conditionsCombustion is often very sensitive to inlet boundary conditionsCorrect velocity and scalar profiles can be critical
Wall heat transfer is challenging to predict; if known, specify wall temperature instead of external convection/radiation BC
DiscretizationStart with first order, then converge with second order to improve accuracy Second order discretization especially important for tri/tet meshes
Initial conditionsWhile steady-state solution is independent of the IC, poor IC may cause divergence due to the number and nonlinearity of the transport equations Cold flow solution, then gas combustion, then particles, then radiation© 2009 ANSYS, Inc. All rights reserved.
8-2
ANSYS, Inc. Proprietary
燃烧;数值模拟
Advanced FLUENT Training Combustion June 2009
Solution strategies for gas phase reactionsNon-premixed model:in general there is no need to solve first the cold flow, or to patch high temperature Start with the reacting flow simulation without radiation Enable radiation once the main flow feature and temperature field have been established
Eddy dissipation/finite rate model:Start with a cold flow solution Patching of products and/or high temperature is needed to start the reactions Use temperature dependent cp’s to avoid unrealistically high temperatures
© 2009 ANSYS, Inc. All rights reserved.
8-3
ANSYS, Inc. Proprietary
燃烧;数值模拟
Advanced FLUENT Training Combustion June 2009
Solution strategies for gas phase reactionsDefault URF could be too aggressive for complex reacting flow system The effect of underrelaxation is highly non-linearDecrease the diverging residual URF in increments of 0.1 Underrelax density when using the mixture-fraction PDF model (0.7) Underrelax velocity for high buoyancy flows Underrelax species to start up the solution (0.9 or lower) Once solution is stable, attempt to increase species, energy, mixture and radiation URF’s as close as possible to 1
Best Practice for the Non-Premixed model are available on the user service center
© 2009 ANSYS, Inc. All rights reserved.
8-4
ANSYS, Inc. Proprietary
燃烧;数值模拟
Advanced FLUENT Training Combustion June 2009
Solution strategies for DPM reacting flow (steady state)Converge the non –reacting flow using first order discretization After the flow field has been established start the particle tracking Depending on the model and conditions patching might be requiredPatching high temperature to start the evaporation or devolatilisation If using the Eddy dissipation model patching some products to start the gas phase reactions
Run the case tracking particle every 20-30 gas phase iterations and lowering the URF for t
he DPM source term (0.1-0.2) Enabled radiation only after the main temperature field and flame shape have been established Solve until a good heat and mass balance have been achieved
© 2009 ANSYS, Inc. All rights reserved.
8-5
ANSYS, Inc. Proprietary
燃烧;数值模拟
Advanced FLUENT Training Combustion June 2009
Solution Strategies for DPM - Unsteady FlowsSet the initial condition for discrete phase The DPM source terms are updated only every particle iteration. If the under-relaxation factor is smaller than one, take care that sufficient particle iterations are performed within the time step in order to achieve the full source terms. You must use a factor of 1, when doing only one particle computation per time step. Lower the Number of Continuous Phase Iterations per DPM Iteration if you need to lower the DPM under-relaxation factor. Use smaller time steps if it does not converge. (make sure that the solution is converge within each time step)
© 2009 ANSYS, Inc. All rights reserved.
8-6
ANSYS, Inc. Proprietary
燃烧;数值模拟
Advanced FLUENT Training Combustion June 2009
Troubleshooting DPMStart DPM particle/droplet too early before the flow field has been developed is often the cause of convergence issues in reacting flow DPM problem Increasing the number of tries and having more iterations between DPM tracks generally helps to make the solution more stable If there are a large number of incomplete particle you should increase the max number steps in the Discrete Phase model panel
© 2009 ANSYS, Inc. All rights reserved.
8-7
ANSYS, Inc. Proprietary
燃烧;数值模拟
Advanced FLUENT Training Combustion June 2009
Solution strategies for detailed chemistryInitial conditions and set upA poor initial condition might cause the stiff chemistry solver to fail A good initial solutions can be calculated using the non-premixed or eddy dissipation models, that would provide a good initial guess (temperature and species) for the stiff chemistry solver