A method capable of simultaneously calculating the pollutants CO and NO in the combustion chamber x numerical values

Abstract

The present invention discloses a method capable of simultaneously calculating pollutants CO and NO in a combustion chamber x The steps are as follows: define a resolution control function; calculate the filtering scale coefficient in the resolution control function, and further calculate the size of the resolution control function; achieve adaptive smooth switching between the unsteady RANS mode, LES mode, and DNS mode in the overall computational domain; remold the turbulent viscosity coefficient in the turbulent model to obtain a new remolded turbulent viscosity coefficient; construct a partially premixed FGM laminar thermodynamics table based on the characteristic variables mixture fraction and reaction progress variable to calculate the CO pollutant; construct a chemical reactor network CRN model, set the global as a standard PSR chemical reactor, and establish an automatic partitioning rule to calculate NO x pollutants. The method of the present invention can achieve high-precision calculation of general combustion chamber pollutants CO and NO x at a lower grid quantity, and predict a more detailed pollutant distribution.

Description

Technical Field

[0001] The present invention belongs to the field of aerospace science and technology and the field of numerical simulation of turbulent flow, and particularly relates to a method capable of simultaneously calculating combustion chamber pollutants CO and NO x numerical values. Background Technique

[0002] With the continuous improvement and strengthening of environmental protection-related laws and regulations, the organization of low-pollution combustion processes has become increasingly important, and the design requirements for low-emission combustion chambers have become more stringent. The traditional design process of combustion chambers has a long cycle and slow performance evaluation. The rapid development of numerical simulation calculation technology provides strong support for economically and efficiently predicting the distribution characteristics of combustion pollutants, and has gradually become an important auxiliary tool for industrial applications.

[0003] The calculation of combustion chamber pollutants is complex and diverse, and a stable, fast and highly accurate turbulent chemical combustion coupling model is required. Currently, turbulent numerical simulation methods are mainly divided into Reynolds-averaged Navier-Stokes (RANS), large eddy simulation (LES), and direct numerical simulation (DNS). The RANS method models the vorticity structures of all scales in the turbulent flow field and performs poorly in the simulation of high Reynolds number unsteady turbulent combustion. Different flows depend on the selection of different RANS models; the LES and DNS methods have high requirements for computing resources, especially for the solution of the near-wall region of high Reynolds number flows, which limits the application of LES and DNS in engineering turbulent combustion problems. The hybrid RANS / LES method has developed rapidly in recent years, and many simulation methods such as PANS and PITM have emerged.

[0004] Two important representatives of combustion pollutants are CO and NO x and the generation mechanisms of the two are significantly different. Usually, the numerical prediction of CO and NO x requires the use of detailed chemical reaction mechanisms, which consume huge computing resources for traditional combustion models such as Eddy Dissipation Concept (EDC). In recent years, the Flamelet Generated Manifold (FGM) combustion model developed by scholars such as Van Oijen, as one of the representative methods for detailed chemical thermodynamics tabulation, uses a finite number of characteristic scalars to describe the chemical reaction process and looks up tables to obtain various physical information in the turbulent combustion field. Even when using detailed chemical reaction mechanisms, since only the transport equations of the characteristic scalars are solved, the occupation of computing resources can be greatly reduced. However, the FGM combustion model is based on a short time scale and is more suitable for the prediction of pollutant CO. For pollutants such as NO x whose generation chemical reactions on long time scales become obvious only after complete combustion, it cannot be correctly described and the calculation accuracy is poor. Currently, there is still a lack of a relatively unified simulation method.

[0005] In summary, for the turbulent combustion pollutant emissions represented by gas turbine combustion chambers, CO and NO xare two important types of pollutants. Currently, most methods have low prediction accuracy for CO and NO x , or can only predict CO. There are few methods that can accurately predict both. Based on this, developing a numerical calculation method that can accurately predict CO and NO x pollutants simultaneously is of great significance. Summary of the Invention

[0006] Aiming at the deficiencies of the above-mentioned existing technologies, the purpose of the present invention is to provide a method for simultaneously calculating the pollutants CO and NO in a combustion chamber x numerically, so as to solve the problems in the existing pollutant calculation process, such as a large time-scale span, low accuracy, high calculation cost, and the ability to only predict CO or NO separately x . The method of the present invention can achieve high-precision calculation of the general combustion chamber pollutants CO and NO at a low grid quantity, and predict a more detailed pollutant distribution. x

[0007] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0008] A method for simultaneously calculating the pollutants CO and NO in a combustion chamber according to the present invention x numerically is as follows:

[0009] (1) Define the resolution control function F r ;

[0010] (2) Calculate the filtering scale coefficient C in the resolution control function F in step (1) through the Smag sub-grid model r , and further calculate the magnitude of the resolution control function F x ; r

[0011] (3) As the discrete grid scale changes, adaptively adjust the resolution control function in step (1) to achieve an adaptive smooth switch between the unsteady RANS mode, LES mode, and DNS mode in the overall calculation domain;

[0012] (4) Based on the resolution control function F r , remold the turbulent viscosity coefficient μ in the Realizable k-ε turbulence model t to obtain a new remolded turbulent viscosity coefficient

[0013] (5) Construct a partially premixed FGM laminar thermodynamics table based on the characteristic variables mixture fraction Z and reaction progress variable C to calculate the CO pollutant;

[0014] ​​(6) Construct a chemical reactor network (CRN) model, set the global standard PSR chemical reactor, establish an automatic partitioning rule to calculate NO x pollutants.

[0015] Furthermore, the specific steps of step (1) are as follows:

[0016] Define the resolution control function F r as:

[0017]

[0018] where is the minimum value function, exp(d) = e d is the natural exponential function; n, β are model empirical constants; L i , L c and L k are the integral length scale, the turbulent cutoff length scale, and the Kolmogorov length scale respectively, and their expressions are:

[0019]

[0020]

[0021]

[0022] In the formula, C x is the filtering scale coefficient; Δ x Δ y Δ z are the scale sizes of the grid in the x, y, and z directions respectively, which are determined by the discretization process; v is the fluid viscosity coefficient; k is the turbulent kinetic energy, and ε is the turbulent dissipation rate.

[0023] Furthermore, the specific steps of step (2) are as follows:

[0024] Calculate the filtering scale C x by the Smagorinsky sub-grid model as:

[0025]

[0026] In the formula, C μ is a turbulent model theoretical constant, and its value is 0.09; C s is a classical coefficient in the Smagorinsky sub-grid model, and the commonly used value in homogeneous shear flow is 0.1. The calculated C x is 0.61.

[0027] Furthermore, the specific steps of step (3) are as follows:

[0028] Define the resolution control function Fr The value is between 0 and 1, and in the computational domain, F r The value is adaptively distributed; when F r The value approaches 1 locally, the RANS model is called for solution; when F r The value approaches 0 locally, the LES / DNS model is called for solution. As it approaches zero, the higher the proportion of the DNS model for solution, and more small-scale vortex structures in the flow field are resolved at this time.

[0029] Further, the specific steps of step (4) are as follows:

[0030] The turbulent viscosity coefficient μ in the Realizable k-ε model t is defined as follows:

[0031] μ t = ρC μ k 2 / ε

[0032] In the formula, ρ is the fluid density;

[0033] For the turbulent viscosity coefficient μ t re-modeling is carried out, and the re-modeled turbulent viscosity coefficient is denoted as Specifically as follows:

[0034]

[0035] Further, the specific steps of step (5) are as follows:

[0036] (51) Use partially premixed FGM laminar thermodynamics to build a table, and build a laminar thermodynamics table based on the characteristic variables mixture fraction Z and reaction progress variable C; use the assumed PDF method to decouple the calculation of flow and chemical reaction, and the assumed probability density functions of the mixture fraction Z and reaction progress variable C are determined by their mean values and variances respectively; in specific calculations, only the mixture fraction and variance reaction progress variable and variance transport equations need to be calculated, and the equations are as follows:

[0037]

[0038]

[0039]

[0040]

[0041] In the formula, the superscript "-" represents spatial filtering, "~" represents Fave filtering; Sc t= 0.7 is the Schmidt number; λ = μC p / Pr is the molecular thermal diffusion coefficient, μ is the dynamic viscosity, C p is the heat capacity coefficient, Pr = 0.85; C g = 2.86, C d = 2.0; is the source term of the reaction progress variable, which is closed by the Zimont model; c φ = 2.0; τ turb = k / ε is the turbulent time scale;

[0042] (52) Using the remodeled and the partially premixed FGM laminar thermodynamics table in the above steps, calculate the vector field and scalar field of the combustion chamber pollutants, obtain the numerical distributions of relevant velocity, temperature, components, and characteristic scalars, and complete the numerical prediction of the combustion-associated pollutant CO.

[0043] Further, the specific steps of step (6) are as follows:

[0044] According to the relevant vector field and scalar field parameters, mark the regions in the burner with similar spatial positions and flame properties, and reaggregate and divide the computational grid into several standard PSR chemical reactors according to the characteristic scalar or vector correlation of the temperature field and mixture fraction; each chemical reactor is connected to each other by mass flow conservation to construct a chemical reactor network CRN model, and the corresponding component concentration values are obtained by solving the component conservation equations inside each chemical reactor to complete the distribution prediction of the post-combustion pollutant NO x ; The calculation formula is:

[0045]

[0046]

[0047] In the formula, represents the mass flow rate, and the subscripts o and p represent the labels of specific chemical reactors, represents the reaction rate of a certain component s in the oth reactor, represents the mass fraction of a certain component s in the oth reactor, represents the mass source term.

[0048] Further, the number and division rules of the chemical reactors in step (6) are specified by the user.

[0049] Advantages of the present invention:

[0050] The method of the present invention calculates the filtering scale coefficient C through the Smag sub-grid model x , and defines the resolution control function Fr achieved a high-precision solution of the flow field in the combustion chamber, coupled with a partially premixed FGM laminar thermodynamics table construction method, to obtain the accurate distribution characteristics of pollutant CO in efficient unsteady calculations. At the same time, coupled with an automatically partitioned CRN chemical reactor network model, it realized the rapid reconstruction of unsteady combustion pollutant calculation data and achieved the accurate prediction of NO x pollutants, providing strong support for the decoupled and efficient calculation of pollutants across time scales in the combustion chambers of many power equipment such as aeroengines and ground gas turbines. Brief Description of the Drawings

[0051] Figure 1 is a flow chart of the method of the present invention.

[0052] Figure 2 is a schematic structural diagram of the Sandia_D type burner in the embodiment of the present invention.

[0053] Figure 3 is a comparison chart of the calculation results of pollutant CO of the burner in the embodiment of the present invention.

[0054] Figure 4 is for pollutant NO in the burner in the embodiment of the present invention x comparison chart of the calculation results. Detailed Embodiments

[0055] For the convenience of those skilled in the art, the present invention will be further described below in conjunction with the embodiments and the drawings. The content mentioned in the embodiments does not limit the present invention.

[0056] Referring to Figure 1 as shown, a method for simultaneously calculating the pollutants CO and NO in the combustion chamber of the present invention x numerical values is as follows:

[0057] (1) Define the resolution control function F r ; The specific steps are as follows:

[0058] Define the resolution control function F r as:

[0059]

[0060] Among them, is the minimum value function, exp(d)=e d is the natural exponential function, n, β are model empirical constants, L i , L c and L k are the integral length scale, the turbulence cutoff length scale and the Kolmogorov length scale respectively, and the expressions are respectively:

[0061]

[0062]

[0063]

[0064] Wherein, C x is the filtering scale coefficient; Δ x Δ y Δ z are the scale sizes of the grid in the x, y, and z directions respectively, determined by the discretization process; v is the fluid viscosity coefficient; k is the turbulent kinetic energy, and ε is the turbulent dissipation rate.

[0065] (2) Calculate the filtering scale coefficient C r in the resolution control function F x in the step (1) through the Smag sub-grid model, and further calculate the size of the resolution control function F r ; The specific steps are as follows:

[0066] Calculate the filtering scale C x through the Smag sub-grid model as:

[0067]

[0068] Wherein, C μ is the theoretical constant of the turbulence model, and its value is 0.09; C s is the classical coefficient in the Smag sub-grid model, and the commonly used value 0.1 in the homogeneous shear flow is adopted, and the calculated C x is 0.61.

[0069] (3) As the discrete grid scale changes, adaptively adjust the resolution control function in the step (1) to achieve an adaptive and smooth switch between the unsteady RANS mode, LES mode, and DNS mode in the overall computational domain; The specific steps are as follows:

[0070] The value of the defined resolution control function F r is between 0 and 1, and the value of F r is adaptively distributed in the computational domain; When the value of F r locally approaches 1, call the RANS mode to solve; When the value of F r locally approaches 0, call the LES / DNS mode to solve. The closer it approaches zero, the higher the solution ratio of the DNS mode, and more fine vortex structures in the flow field are resolved at this time.

[0071] (4) Based on the resolution control function F r, remold the turbulent viscosity coefficient μ in the Realizable k-ε turbulence model t , and obtain the new remolded turbulent viscosity coefficient The specific steps are as follows:

[0072] The turbulent viscosity coefficient μ in the Realizable k-ε model t is defined as follows:

[0073] μ t = ρC μ k 2 / ε

[0074] where ρ is the fluid density;

[0075] Remold the turbulent viscosity coefficient μ t , and the remolded turbulent viscosity coefficient is denoted as Specifically as follows:

[0076]

[0077] (5) Construct a partially premixed FGM laminar thermodynamics table based on the characteristic variables mixture fraction Z and reaction progress variable C to calculate CO pollutants; the specific steps are as follows:

[0078] (51) Use the partially premixed FGM laminar thermodynamics to build a table, and construct a laminar thermodynamics table based on the characteristic variables mixture fraction Z and reaction progress variable C; use the assumed PDF method to decouple the calculation of flow and chemical reactions, and the assumed probability density functions of the mixture fraction Z and reaction progress variable C are determined by their mean values and variances respectively; in specific calculations, only the mixture fraction and variance reaction progress variable and variance transport equations need to be calculated, and the equations are as follows:

[0079]

[0080]

[0081]

[0082]

[0083] where the superscript "-" represents spatial filtering, and "~" represents Fave filtering; Sc t = 0.7 is the Schmidt number; λ = μC p / Pr is the molecular thermal diffusion coefficient, μ is the dynamic viscosity, C p is the heat capacity coefficient, Pr = 0.85; Cg = 2.86, C d = 2.0; is the source term of the reaction progress variable, closed by the Zimont model; c φ = 2.0; τ turb = k / ε is the turbulent time scale;

[0084] (52) Use the remodeled and the partially premixed FGM laminar thermodynamics table in the above steps to calculate the vector field and scalar field of the combustion chamber pollutants, obtain the numerical distributions of relevant velocities, temperatures, components, and characteristic scalars, and complete the numerical prediction of the combustion-associated pollutant CO.

[0085] (6) Construct a chemical reactor network CRN model, set the global as a standard PSR chemical reactor, establish an automatic partitioning rule to calculate NO x pollutants; the specific steps are as follows:

[0086] According to the relevant vector field and scalar field parameters, mark the regions in the burner with similar spatial positions and flame properties, and reaggregate and divide the computational grid into several standard PSR chemical reactors according to the characteristic scalar or vector correlation of the temperature field and mixture fraction; each chemical reactor is connected to each other by mass flow conservation to construct a chemical reactor network CRN model, and the corresponding component concentration values are obtained by solving the component conservation equations inside each chemical reactor to complete the distribution prediction of the post-combustion pollutant NO x ; the calculation formula is:

[0087]

[0088]

[0089] In the formula, represents the mass flow rate, and the subscripts o, p represent the labels of specific chemical reactors, represents the reaction rate of a certain component s in the oth reactor, represents the mass fraction of a certain component s in the oth reactor, represents the mass source term.

[0090] Among them, the number of chemical reactors and the partitioning rule in step (6) are specified by the user.

[0091] Example 1:

[0092] See Figure 2 shown, the Sandia_D type burner is used to verify an invention of the present invention for simultaneously and accurately calculating the combustion chamber pollutants CO and NO xThe numerical calculation method uses hexahedral meshes to discretize the computational domain and compares the simulation results with those of the traditional RANS-FGM combustion model for verification. The fuel stage is the main fuel inlet, and the fuel composition is specifically a methane premixed gas with a volume ratio of 25%:75%. The value class flame inlet is composed of the admixture of acetylene, hydrogen, air, etc., and its mixture is a burned methane / air premixed gas with an enthalpy equivalent ratio of 0.77. The coflow is air. In the CFD model, the pressure-velocity coupling algorithm uses SIMPLEC. The time term uses a bounded second-order central implicit advancement; the momentum flux is discretized using a bounded second-order central scheme; and the scalar fluxes are all discretized using a second-order upwind scheme.

[0093] Under the framework of the finite volume method CFD software, according to the steps shown above, the code of the model method of the present invention is implanted and compiled and executed, and the simulation results are compared with the experimental results; the distributions of CO and NO as shown in Figure 3 and Figure 4 are respectively compared. Under the method of the present invention, both the prediction of the CO distribution peak value and the inflection point are in good agreement with the experimental values, especially the prediction of the pulsation value has a high accuracy; the NO x prediction results are in agreement with the experimental distribution, not only more accurate numerically, but also very similar in the distribution trend along the axis. x There are many specific application ways of the present invention. The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements can be made, and these improvements should also be regarded as the protection scope of the present invention.

[0094] There are many specific application ways of the present invention. The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements can be made, and these improvements should also be regarded as the protection scope of the present invention.

Claims

1. A method capable of simultaneously calculating the pollutants CO and NO in the combustion chamber x values It is characterized in that The steps are as follows: (1) Define the resolution control function F r ; (2) Calculate the filtering scale coefficient C in the resolution control function F in the step (1) through the Smagorinsky sub-grid model r and further calculate the magnitude of the resolution control function F x ; r ​ (3) As the discrete grid scale changes, adaptively adjust the resolution control function in the step (1) to achieve an adaptive and smooth switch between the unsteady RANS mode, LES mode, and DNS mode in the overall computational domain; (4) Based on the resolution control function F r , re-model the turbulent viscosity coefficient μ in the Realizable k-ε turbulence model t , and obtain the new modeled turbulent viscosity coefficient (5) Construct a partially premixed FGM laminar thermodynamic table based on the characteristic variable mixture fraction Z and the reaction progress variable C to calculate CO pollutants; (6) Construct a chemical reactor network (CRN) model, set the global to a standard PSR chemical reactor, establish an automatic partitioning rule to calculate NO x pollutants; The specific steps of the step (1) are as follows: Define the resolution control function F r as follows: Among them, is the minimum value function, and exp(d) = e d is the natural exponential function; n and β are model empirical constants; L i , L c and L k are the integral length scale, the turbulence cutoff length scale, and the Kolmogorov length scale respectively, and their expressions are respectively: where C x is the filtering scale coefficient; Δ x Δ y Δ z are the scale sizes of the grid in the x, y, and z directions, respectively, determined by the discretization process; v is the fluid viscosity coefficient; k is the turbulent kinetic energy, and ε is the turbulent dissipation rate.

2. The method according to claim 1 for simultaneously calculating the values of pollutants CO and NO in the combustion chamber x of the combustion chamber It is characterized in that The specific steps of the step (2) are as follows: Calculating the filtering scale C through the Smagorinsky sub-grid model x is as follows: In the formula, C μ is the theoretical constant of the turbulence model, and its value is 0.09; C s is the classical coefficient in the Smagorinsky sub-grid model, and the commonly used value 0.1 in homogeneous shear flow is adopted. The calculated C x is 0.

61.

3. The method according to claim 2 for simultaneously calculating the pollutants CO and NO in the combustion chamber x values It is characterized in that The specific steps of the step (3) are as follows: The defined resolution control function F r has a value between 0 and 1, and in the computational domain, F r is adaptively distributed; when F r locally approaches 1, the RANS mode is called for solution; when F r locally approaches 0, the LES / DNS mode is called for solution. The closer it approaches zero, the higher the proportion of the DNS mode in the solution, and at this time, more small vortex structures in the flow field are resolved.

4. The method according to claim 3 for simultaneously calculating the pollutants CO and NO in the combustion chamber x values It is characterized in that The specific steps of the step (4) are as follows: Turbulent viscosity coefficient μ in the realizable k-ε model t It is defined as follows: μ t = ρC μ k 2 / ε In the formula, ρ is the fluid density; For the turbulent viscosity coefficient μ t Re-model it, and the re-modeled turbulent viscosity coefficient is denoted as Specifically as follows:

5. The method according to claim 4 for simultaneously calculating the values of pollutants CO and NO in the combustion chamber x of the combustion chamber It is characterized in that The specific steps of the step (5) are as follows: (51) Use the partially premixed FGM laminar thermodynamics to build a table, and construct a laminar thermodynamics table based on the characteristic variables mixture fraction Z and reaction progress variable C; use the assumed PDF method to achieve decoupled calculations of flow and chemical reactions, and the assumed probability density functions of the mixture fraction Z and reaction progress variable C are determined by their mean values and variances respectively; in specific calculations, only the mixture fraction and variance reaction progress variable and its variance transport equations, and the equations are shown as follows: In the formula, the superscript "—" represents spatial filtering, and "~" represents Fave filtering; Sc t = 0.7 is the Schmidt number; λ = μC p / Pr is the molecular thermal diffusion coefficient, μ is the dynamic viscosity, C p is the heat capacity coefficient, Pr = 0.85; C g = 2.86, C d = 2.0; is the source term of the reaction progress variable, which is closed by the Zimont model; c φ = 2.0; τ turb = k / ε is the turbulent time scale; (52) Using the remodeled one in the above steps and the partial premixed FGM laminar thermodynamics table, perform vector field and scalar field calculations on the combustion chamber pollutants to obtain the numerical distributions of relevant velocities, temperatures, components, and characteristic scalars, and complete the numerical prediction of the combustion-associated pollutant CO.

6. The method according to claim 5 for simultaneously calculating the pollutants CO and NO in the combustion chamber x values It is characterized in that The specific steps of the step (6) are as follows: According to relevant vector field and scalar field parameters, regions with similar spatial positions and flame properties in the burner are marked, and the computational grid is re-aggregated and divided into several standard PSR chemical reactors according to the characteristic scalar or vector correlation of the temperature field and mixture fraction; each chemical reactor is connected to each other through mass flow conservation to construct a chemical reactor network CRN model, and the corresponding component concentration values are obtained by solving the component conservation equations inside each chemical reactor to complete the distribution prediction of pollutants NO x after combustion; The calculation formula is: In the formula, represents the mass flow rate, and the subscripts o and p represent the labels of specific chemical reactors. represents the reaction rate of component s in the o-th reactor. represents the mass fraction of component s in the o-th reactor. represents the mass source term.

Images