CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

CFX-5 Solver and Solver Manager

Combustion in CFX-5
Introduction p. 444 The Eddy Dissipation Combustion Model. p. 445 Theory Documentation p. 448 Phasic Combustion p. 455

Page 443

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

Introduction
CFX-5 includes combustion models to allow the simulation of flows in which combustion reactions occur. Three models are currently available, the eddy dissipation model (EDM), a finite rate chemistry model(FRC) and a combined EDM/FRC model. The eddy dissipation model due to Magnussen and Hjertager [38] was developed for use in a wide range of turbulent reacting flows covering premixed and diffusion flames. Because of its simplicity and robust performance in predicting turbulent reacting flows, this model has been widely applied in the prediction of industrial flames. The finite rate chemistry model allows the computation of reaction rates described by the molecular interaction between the components in the fluid. It can be combined with the eddy dissipation model for flames where chemical reaction rates might be slow compared with the reactant mixing rates. It also allows the use of user customised reaction rate expressions via CFX Expression Language (CEL). These models determine the rates at which a component is consumed or created in a single reaction step during the combustion process. For multiple step reactions, the reactions are added to obtain the total reaction rate. In order to enhance steady convergence and allow the use of a larger time step, CFX-5 uses a coupled linear solver for the transport equations of the chemically active components. In the CFX-5 Solver, no limit has been set for the maximum number of reaction steps or components in a fluid. However, the larger the number of components, the larger will be the memory requirements. The first section of the documentation shows how to use the different combustion models in CFX-5. A description of the theory of the eddy dissipation and finite rate chemistry models is provided in the second section. This chapter extends the ideas covered in Multicomponent Flow (p. 367). You should be familiar with the multicomponent flow chapter before reading this section.

Page 444

Introduction

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

The Eddy Dissipation Combustion Model.

To use the model in CFX-5, you must first set up a reacting fluid. In the Domains panel, after setting your Domain Options, click on Fluids Editor. Select Create->Fluid->Reacting Mixture. For CFX-5.5.1, the only reactions that can be specified from the GUI are the predefined reactions that appear in the Reactions... list. It is possible to create your own reactions outside of CFX-Build and have them available when CFX-Build is next loaded. In <CFXROOT>/ etc/5.5.1 there are two files named MATERIALS and REACTIONS. Each of these files contain instructions for creating further materials and reactions. This is an advanced feature for this release, and full documentation will be available in the beta feature binder and future releases of CFX-5. Thermodynamic data for the MATERIALS file was obtained from[44]. If you wish to specify any non-reacting components that will also be present in the domain, you can enable the check mark next to Include Passive Components and then select each one from the Passive Components... list. Keep in mind that the thermodynamic properties for materials from the CFX-Build database may not be appropriate (constant specific heat capacity). This may produce poor results if a passive component has a significant mass fraction. Use materials from the MATERIALS database if possible. Finally, enter a name for the fluid in the Name box and click Apply. Click Cancel to close the form and return to the Domains form. On the Fluid Models form the option Combustion Model is now available. The default is the Eddy Dissipation model, which should be left unchanged. Click on the Components button to open the Component Models form. On this form you can select the type of equation used for each component. Clicking on Equation Type lets you select between a Transport Equation, Algebraic Equation or Constraint. If you wish to include a radiation model with your calculation, you can do so by choosing the P 1 or Rosseland model from the Radiation Model menu entry. The default is No Radiation. For further information on the radiation models used in CFX-5, please see Radiation in CFX-5 (p. 459).

Page 445

The Eddy Dissipation Combustion Model.

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

The Eddy Dissipation model requires that fuel, oxidant and products be available in a control volume before combustion will occur. However, combustion products may not always be specified as an input. In this case, products could not form unless they are introduced into the domain. Assuming the problem is one in which a stable flame may be established, initial specification of products within the domain should be sufficient to start and maintain combustion. However, if the combustion is difficult to maintain, it may be necessary to introduce a small fraction of products in one of the inlets. To specify the initial conditions click on the Init. Cond, button and select Set Individually for the Variables button.Click on Variables and then on Components to set the mass fraction for each of the components. After setting these you can set the remainder of the initial conditions for the domain.

Troubleshooting
The combustion model has been designed to obtain a robust, accurate solution of the combustion source terms to the scalar and energy equations. However, in some cases difficulties may arise in trying to obtain a converged solution. Combustion introduces a strong coupling between scalars, energy and momentum. In difficult problems, you may find it helps to calculate a cold flow solution before turning on combustion. The initial guess for scalars can have a large effect on the convergence history of a combustion problem. For example, initial specification of a large concentration of fuel and oxidant in the domain can produce high, unrealistic temperatures that slow convergence. It is best to specify the concentrations in such a way that the reaction proceeds starting from the inlet and proceeds through the domain. This will help ensure reasonable flame temperatures during convergence. Time step choice may also have a significant effect on the convergence of a flow calculation. Combustion introduces large density gradients in the flow field which can increase the sensitivity of the continuity equation to the time step. If you are having difficulty converging a combusting flow, particularly during an early stage of convergence, try reducing the time step to improve solution of the continuity equation. For further advice, please consult Advice for modelling Combusting Flows in CFX-5 (p. 456).
Page 446

The Eddy Dissipation Combustion Model.

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

If the solution of a combusting flow appears to stall with mass not well conserved, use a time step appropriate to the fluid flow for a few iterations to push fluid through the system. In general for the Finite Chemistry Model, the time scale of the combustion process is smaller, so you may have to continue with a smaller time step to force the mass fraction and energy equations to converge. The timescale for the Eddy Dissipation Model is equal to the turbulence time scale ( k ). In some cases, it may be appropriate to use a larger timescale for the mass fraction equations. This is particularly true when there are regions of small flow velocity in the domain. In these cases, the time needed to convect the components into all areas of the domain may be relatively large.

Page 447

The Eddy Dissipation Combustion Model.

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

Theory Documentation
This section outlines the basis of the implementation of combustion modelling in CFX-5.5.1. First, the transport equations for energy and the components are revisited, then the chemical reaction rate computation is described, and finally, the computation of the rate of progress of a chemical reaction is explained in the context of the Eddy Dissipation and Finite Rate Chemistry Models. Extinction is modelled by setting the reaction rate locally to zero.

Transport Equations
Combustion models in CFX-5.5.1 use the same algorithm used for Multicomponent Fluid with the addition of a source/sink term due to chemical reactions. The equation of transport for component I with mass fraction, YI is then ( Y I ) ( u j Y I ) Y I --------------- + ---------------------- = ------- I eff -------- + S I x j t x j x j

(Eq. Sol.-77)

where the source term SI is due to the chemical reaction rate involving component I.

Chemical Reaction Rate

In general, chemical reactions can be described in terms of K elementary reactions involving NC components that can be written as
kI I = A, B, C, NC NC

I = A, B, C,

kI I

(Eq. Sol.-78)

where vkI is the stoichiometric coefficient for component I in the elementary reaction k. The rate of production/consumption, SI, for component I can be computed as the sum of the rate of progress for all the elementary reactions that component I participates in:

Page 448

Theory Documentation

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

:S I = W I

k=1

( kI kI )R k

(Eq. Sol.-79)

where Rk is the elementary reaction rate of progress for reaction k, which in CFX-5.5.1 can be calculated using the Eddy Dissipation Model or/and the Finite Rate Chemistry Model.

Fluid Timescale for Extinction Model

When the model for flame extinction at high turbulence is activated, local extinction occurs when the turbulence timescale is smaller than a chemical timescale (quenching timescale) provided by the user i.e. whenever t < c . While c is directly specified , the turbulence timescale is computed from the CFD solution fields. One possibility is to apply the Kolmogorov timescale, Kolmogorov = v

An alternative is to use the mixing timescale, k mixing = As this is a very simple model for predicting local extinction, the specified chemical timescale may need to be adjusted in order to achieve best results for a specific problem. For methane-air combustion, good starting points are 1.37e-4 [s] when applying the Kolmogorov timescale, or 5e-4 [s] when comparing to the mixing timescale. Using the Kolmogorov timescale tends to be more aggressive and may lead to global extinction of the flame, even in situations where this is not physical. It is for this reason that the mixing timescale is recommended. By default the mixing timescale is applied. This may be changed by setting the expert parameter use kolmogorov ts for extinction to T.

Page 449

Theory Documentation

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

The Eddy Dissipation Model

The eddy dissipation model is based on the concept that chemical reaction is fast relative to the transport processes in the flow. When reactants mix at the molecular level, they instantaneously form products. The model assumes that the reaction rate may be related directly to the time required to mix reactants at the molecular level. In turbulent flows, this mixing time is dominated by the eddy properties, and therefore, the rate is proportional to a mixing time defined by the turbulent kinetic energy, k, and dissipation, . rate k This concept of reaction control is applicable in many industrial combustion problems where reaction rates are fast compared to reactant mixing rates. In summary: The Eddy Dissipation model is only applicable in turbulent flows. There is no kinetic control of the reaction process. Thus, ignition and processes where chemical kinetics may limit reaction rate may be poorly predicted. In the Eddy Dissipation model, the rate of progress of elementary reaction k, is determined by the smallest of the two following expressions: Reactants Limiter [I ] -min ------- Rk = A k v kI where [ I ] is the molar concentration of component I and I only includes the reactant components.

Page 450

Theory Documentation

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

Products Limiter [ I ]W I P R k = AB - ----------------------- k W P kI I where P loops over all product components in the elementary reaction k. The products limiter may be disabled by setting the model coefficient B to a negative value. For multi-step reaction schemes it is turned off by default, but may be turned on by explicitly setting the model coefficient B to a positive value (although this is not recommended).

The Finite Rate Chemistry Model

The Finite Rate Chemistry model, as implemented in CFX-5.5.1, assumes that the rate of progress of elementary reaction k can be reversible only if a backward reaction is defined. Therefore, the rate of progress Rk, is computed as
C C r r kI Rk = F k [ I ] B k [ I ] kI I = A, B, I = A, B,

where [ I ] is the molar concentration of component I and Fk and Bk are the forward and backward rate constants respectively. r represent the reaction order of component I in the elementary reaction k. This reaction order is equal to the stoichiometric coefficient for elementary reactions, but it can be different for certain global reactions. The only built-in formula for the forward and backward rate constants assumes an Arrhenius temperature dependence as

= A T k k

Ek exp ------- RT Ek exp ------- RT

CFX-5.5.1

B = A T k k
Page 451

Theory Documentation

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

where: Ak is pre-exponential factor k is the temperature exponent (dimensionless) Ek is the activation energy T is the absolute temperature R k can also be specified directly without using relations. Separate sets of coefficients A k , k and E k are applied to forward and backward rates. In cases that a third body is needed for the reaction to occur, the rate of progress described earlier is scaled by

ki [ I i ]
i=1

where ki is the relative participation of component Ii in reaction k. For those components that have a high probability to participate in the reaction, the coefficient ki is higher than those that rarely participate or do not participate at all (ki = 0). CFX-5.5.1 assumes that if a third body is present in the reaction, all components have the same weight, ki = 1. These efficiencies can be specified in the command file for the given reaction.

Combustion Source Term Linearisation

The combustion source terms may have a dominant influence on the solution of the scalar and energy equations. Thus, it is important to treat the combustion source terms carefully in order to obtain robust convergence of the fluid flow.

Page 452

Theory Documentation

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

A property of multicomponent fluids is that the mass fraction of any given component is bounded between 0 and 1. Combustion tends to drive reactant concentrations towards the lower limit and product concentrations toward the upper limit. If the time step is large, the combustion sources may cause scalars to exceed these bounds. Thus, the sources may need to be moderated to maintain physically realistic mass fractions. The combustion sources in CFX-5 have been linearised to prevent the formation of negative mass fractions. Consider the solution of component I with the source term, R, which is calculated from (Eq. Sol.-80) in each control volume. To prevent the mass fraction of I from exceeding the bounds of 0 to 1, the source term is calculated according to: RI + RI RI RI source = ------------------- I + ------------------------- ( 1 I ) 2I 2 ( 1 I )** where I * = max ( , I ) and where (1 - I )** = max ( 1 I ) . and is a small number (set to 10-6). The combustion reaction rate should approach 0 whenever any of the reactant or product mass fractions approach 0. If the source is positive (for products), then the first term on the right hand side of (Eq. Sol.-80) is zero and the source is: source = R I otherwise (1 I ) source = R I -------------- Thus as products 1 , source 0 . If the source is negative (reactants), then the second term on the right hand side of (Eq. Sol.-80) is zero and the source is: source = R I otherwise ( RI I ) source = ------------
Page 453

(Eq. Sol.-80)

if

I1

if

Theory Documentation

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

Thus as reactant 0 , source 0 . This treatment of combustion sources allows larger timesteps to be used in calculating a steady state solution than would be possible without the linearisation.

The Combined Eddy Dissipation/Finite Chemistry Model

For the combined Finite Rate Chemistry/Eddy Dissipation Model, the reaction rates are first computed for each model separately and then the minimum of the two is used. This procedure is applied for each reaction step separately, so while the rate for one step may be limited by the chemical kinetics, some other step might be limited by turbulent mixing at the same time and physical location. It is also possible to apply different combustion models to each of the steps in a multi-step scheme. Some of the predefined schemes make use of this feature, regardless of the global model selection. NO formation steps are always Finite Rate Chemistry and Eddy Dissipation, and the Hydrogen Oxygen step is set to Eddy Dissipation due to the fast chemistry involved in this reaction. The combined model is valid for a wide range of configurations, provided the flow is turbulent. In particular, the model is valid many reactions which range from low to high Damkhler numbers (chemistry slow/fast compared to turbulent timescale). Use of this model is recommended if reaction rates are limited by turbulent mixing in one area of the domain and limited by kinetics somewhere else. The Eddy Dissipation model can, however, be more robust than Finite Rate Chemistry or the combined model.

Page 454

Theory Documentation

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

Phasic Combustion
For multiphase simulations, phasic combustion describes combustion within each individual phase. The combustion models are solved in exactly the same way as for a single phase simulation. Combustion can occur in any phase, and different combustion models can be used in each separate phase. When using multicomponent multiphase flow, a maximum of two fluids can be specified. There is no mass transfer between phases for this release of CFX-5. For details of the models available, please refer to the previous sections The Eddy Dissipation Model (p. 450), The Finite Rate Chemistry Model (p. 451) and The Combined Eddy Dissipation/ Finite Chemistry Model (p. 454).

Page 455

Phasic Combustion

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

Advice for modelling Combusting Flows in CFX-5

Smart Start Smart Start (Global Dynamic Model Control) for combustion disables the product limiter in the Eddy Dissipation and Finite Rate Chemistry/ Eddy Dissipation Models for the first five timesteps. This allows the specification of no products in the domain as a valid initial condition. Global Dynamic Control is always turned on for combustion regardless of the choice selected in CFX-Build, as it provides the most appropriate solver behaviour. You can disable this feature, or change the number of timesteps that the product limiter is disabled for by editing the ccl of your definition file and adding the parameter Transition Iteration.
Note: This feature is only for expert users of the software

For details on editing ccl in definition files, please see Editing the Command Language (CCL) File (p. 84). The resulting CCL would appear as follows: DYNAMIC MODEL CONTROL: Global Dynamic Model Control = No COMBUSTION CONTROL: Transition Iteration = <Integer> END END The <Integer> should be set to zero to turn off global dynamic model control. General Procedure for running more complex simulations For many cases, such as the Combustion Tutorial (Combustion and Radiation in a Can Combustor (p. 501 in CFX-5 Tutorials)), the run can be started using automatic initial guesses for most fields and the choice of model, (EDM) is the most robust of the combustion models. If you are having trouble getting a complex problem to converge, the following steps may aid convergence: Begin by specifying a single step reaction using the Eddy Dissipation Model.
Page 456

Advice for modelling Combusting Flows in CFX-5

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

For initial conditions, specify Oxygen with a mass fraction of 0.232 and set each of the other species to Automatic. Allow the problem to converge to a reasonable level. Restart the problem with Finite Rate Chemistry or Finite Rate Chemistry/Eddy Dissipation Model using the results from the previous problem as an initial guess. Advantages and Disadvantages of using Multistep Reaction Mechanisms Multi-step schemes allow for predicting a larger number of species and intermediate species like CO. They also apply if the fuel is a mixture of several species, e.g. when modelling natural gas as a mixture of CH4, CO and H2. For multi-step schemes the products limiter in the Eddy Dissipation model will be turned off by default (global extinction of the flame would happen if turned on with default EDM coefficients). Because of this, when applying multi-step schemes to premixed or partially premixed systems it is recommended to use the combined FRC/EDM model. It is important to keep in mind that Finite Rate Chemistry does not account for turbulent fluctuations of temperature when computing temperature dependent reaction rates. Another consideration is that even though several reactions steps can be modelled, this usually remains a strong simplification from reality where dozens of species may be involved in hundreds of reactions simultaneously. Sensible Values for Initial Temperature and Mass Fractions For the Eddy Dissipation model, you can safely choose Automatic as the option in most cases. The Finite Rate Chemistry model initial temperature depends on the temperature dependency of the reactions. If no sensible value is known, restarting from a run which used the EDM model can be useful. For any of the predefined reactions, using 0.232 as an initial guess for the mass fraction of Oxygen is a good initial guess. Use Automatic for other species. Another useful initial guess method is to specify according to the mixture stoichiometry (mass flow weighted average of the inflow boundary conditions).
Page 457

Advice for modelling Combusting Flows in CFX-5

CFX-5.5.1

CFX-5 Solver and Solver Manager

Master Contents

Master Index

Help On Help

Retrace

Tips for improving convergence It is often useful to apply a larger time step to temperature and species than the remaining equations. Try starting with a relatively large timestep to let the material sweep through the domain. You can reduce the timestep later to achieve better resolution of the simulation conditions, and therefore convergence. When using the Finite Rate Chemistry model, using a timestep that is too large at the beginning may cause the solver to fail due to reaction rates not converging. Under these circumstances, the timestep must be reduced. Enabling the expert parameter monitor ranges allows you to monitor species mass fractions and temperature. For more information on expert parameters, please see CFX-5 Solver Expert Control Parameters (p. 528). Appropriate Conditions for the Combustion Models

Eddy Dissipation Model

Turbulent Fast reaction compared to turbulent timescale (high Damkhler number) Reaction rate dominated by turbulent mixing of reactants or fresh and burned gases (premixed) Laminar or turbulent If turbulent, reaction rate slow compared to turbulent timescale (low Damkhler number) Reaction rate dominated by kinetics (chemistry) Kinetic data for reaction rates required May need special initialisation for flame ignition (temperature dependence of reaction rates) Turbulent Whole range of Damkhler numbers Kinetic data for reaction rates required May need special initialisation for flame ignition (temperature dependence of reaction rates).

Finite Rate Chemistry Model

Combined EDM/ FRC Model

Page 458

Advice for modelling Combusting Flows in CFX-5

CFX-5.5.1