A method for preserving spatial and temporal correlation scalars based on dual time stepping

By adopting a two-time-step propulsion-based scalar method in the numerical simulation of chemical reaction flow, the problem of negative value of scalar physical quantities in the iteration process is solved, and efficient and stable calculation is achieved The process improves calculation accuracy and efficiency.

CN118797822BActive Publication Date: 2024-11-15CALCULATION AERODYNAMICS INST CHINA AERODYNAMICS RES & DEV CENT
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411288466.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-14
Publication Date
2024-11-15
Estimated Expiration
2044-09-14

AI Technical Summary

Technical Problem

In the numerical simulation of chemical reaction flow, different chemical reaction time scales are inconsistent, resulting in negative scalar physical quantities that may appear during the iteration process, hindering the progress of calculations. The existing high-order dual-time step method is complex and has a computational cost. high.

Method used

The spatial and temporal correlation scalar positive method based on two-time step advancement is adopted. Through the constraints of pseudo-time step length and the calculation of mixed parameters, the scalar positiveness is ensured, and a simple flux restriction method is provided to maintain calculation efficiency and high precision.

Benefits of technology

It realizes efficient scalar positiveness, reduces calculation costs, improves calculation efficiency, and maintains the stability of second-order and third-order accuracy in the Runge-Kutta time propulsion method.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118797822B_ABST
    Figure CN118797822B_ABST
Patent Text Reader

Abstract

The invention discloses a time-space related scalar preservation method based on dual time step advancement, which relates to the fields of aerodynamics and computational fluid dynamics, and comprises: starting fluid mechanics numerical simulation, performing numerical discretization processing, and calculating and obtaining the physical flux of the unit interface; performing feature projection on the physical flux to obtain the conserved variable, and obtaining the diagonal matrix based on the conserved variable calculation; constructing the condition of scalar preservation; judging whether the condition of scalar preservation is met based on the pseudo time step and the diagonal matrix; if the condition is met, performing numerical iterative calculation based on the numerical flux and the corresponding parameters; if not, solving and obtaining the mixed parameter based on the pseudo time step and the diagonal matrix, and obtaining the preserved numerical flux based on the mixed parameter and the numerical flux; performing numerical iterative calculation based on the preserved numerical flux and the corresponding parameters. The invention has high calculation efficiency and high calculation accuracy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the fields of aerodynamics and computational fluid dynamics, and in particular to a time-space related scalar preservation method based on double time step advancement. Background Art

[0002] When a high-speed aircraft flies at a low altitude and high speed, the high-speed incoming flow generates thousands of degrees of temperature through shock wave compression and wall friction, forming high enthalpy conditions, which excites thermodynamic relaxation (vibration excitation temperature is about 800K) and dissociation chemical reactions (oxygen dissociation temperature is about 2000K and nitrogen dissociation temperature is about 4000K) in the air, forming a thermochemical non-equilibrium effect and significantly affecting the flow characteristics in the turbulent boundary layer, which brings great challenges to the accurate evaluation of the aerodynamic thermal characteristics of the aircraft surface. At present, numerical simulation is one of the most important research methods, but the time scales of different chemical reactions in chemical reaction flows are inconsistent. In the iterative process of solving chemical reaction flows, negative scalar physical quantities (such as density, pressure, etc.) may be interpolated. These negative values ​​are non-physical, which seriously hinders the numerical solution and ultimately leads to calculation failure. Therefore, how to maintain the calculation accuracy in numerical calculations while ensuring that the scalar physical quantities in the iterative process are always positive is one of the difficult problems that must be solved in the numerical simulation of chemical reaction flows.

[0003] At present, there has been some progress in the research on time-marching correction methods. For example, scalar correction methods have been proposed in the finite volume method WENO format, the finite difference method WENO format, and the discontinuous finite element method. However, these methods are relatively complex to implement, and have many restrictions on the time step. In addition, the process of converting high-order fluxes to correct fluxes will bring a lot of computational work, resulting in increased computational costs and significantly reduced computational efficiency.

[0004] Therefore, there is still a lack of effective and simple methods to effectively maintain the correctness of high-order dual time-step methods (especially methods with accuracy exceeding second order). Summary of the invention

[0005] The purpose of the present invention is to efficiently realize time-space related scalar correction.

[0006] To achieve the above-mentioned object of the invention, the present invention provides a time-space correlation scalar correction method based on dual time step advancement, the method comprising:

[0007] Step 1: Start the fluid mechanics numerical simulation, perform numerical discretization, and calculate the physical flux at the unit interface i-1. , the physical flux at the unit interface i and the physical flux at the cell interface i+1 Reconstruct the numerical flux at the unit interface i+1 / 2 ,in,i is the ith spatial discrete node, m For the m Take a walk at a time;

[0008] Step 2: Set the numerical flux at the cell interface i+1 / 2 The conserved variables corresponding to the spatial discrete nodes i and i+1 are and Project upward and calculate the diagonal matrices corresponding to the spatial discrete nodes i and i+1 respectively and ,Will Split to obtain the positive eigenvalue diagonal matrix corresponding to the spatial discrete node i and negative eigenvalue diagonal matrix ,Will Split to obtain the positive eigenvalue diagonal matrix corresponding to the spatial discrete node i+1 and negative eigenvalue diagonal matrix ;

[0009] Step 3: Construct the conditions for scalar preservation;

[0010] Step 4: Obtain the pseudo time step Δτ, based on the pseudo time step Δτ, and , determine whether the scalar preservation condition is met;

[0011] If the scalar preservation condition is met, then based on the numerical flux And the corresponding parameters are numerically iterated;

[0012] If the scalar preservation condition is not met, then based on the pseudo time step Δτ, and , solve to obtain the mixing parameter θ, based on the mixing parameter θ and the numerical flux , and obtain the numerical flux of the positive ; then based on the numerical flux of the positive And the corresponding parameters are numerically iterated.

[0013] Among them, based on the positive coefficient rule, the present invention proposes a positive-preserving dual time-step method, which constrains the time iteration step to ensure the positive-preserving characteristics in the dual time-step advancement process, and provides a simple flux limitation method to ensure that the positive flux is calculated. It has high computational efficiency and can achieve scalar positiveness in the Runge-Kutta time advancement method with strong stability of second-order and third-order accuracy.

[0014] Among them, the condition for scalar preservation is:

[0015] ;

[0016] Where Δt is the physical time step, a x for x Characteristic coefficient in direction, Δ ξ is the curve coordinate along ξ The distance of advancement in the direction, I is the unit diagonal matrix.

[0017] in, a x The calculation method is:

[0018] ;

[0019] Among them, τ x for x The maximum characteristic wave speed in the direction, τ y for y The maximum characteristic wave speed in the direction;

[0020] τ x The calculation method is:

[0021] ;

[0022] τ y The calculation method is:

[0023] ;

[0024] in, u 1 and u 2 are x and y The velocity component in the direction, c is the speed of sound; ξ x , ξ y , η x and η y are coordinate transformation coefficients, and is the transformation coefficient, , ; Dξ and D η is the node spacing in the computational coordinate system.

[0025] Among them, the pseudo time step Δτ is calculated as:

[0026] ;

[0027] Among them, τ x for x The maximum characteristic wave speed in the direction, τ yfor y The maximum characteristic wave velocity in the direction, Δt is the physical time step, CFL is the CFL number, which is the basic setting parameter in computational fluid dynamics and the criterion for format stability and convergence, and is generally less than or equal to 1.

[0028] Among them, the calculation method of the mixing parameter θ is:

[0029] ;

[0030] Where Δt is the physical time step, The physical flux Projection on The diagonal matrix after For The diagonal matrix composed of positive eigenvalues ​​obtained by Steger-Warming splitting, For The diagonal matrix composed of negative eigenvalues ​​obtained by Steger-Warming splitting, The physical flux Projection on The diagonal matrix after .

[0031] in, and The calculation method is:

[0032] ;

[0033] in, L i is the eigenvector matrix corresponding to the spatial discrete node i, for L i The inverse matrix of L i+1 is the eigenvector matrix corresponding to the spatial discrete node i+1, for L i+1 The inverse matrix of

[0034] and The calculation method is:

[0035] ;

[0036] ;

[0037] The square brackets [ ] contain matrices, and the subscript r represents the rth element in the matrix, It represents No. r elements.

[0038] Among them, the numerical flux of the positive The calculation method is:

[0039] The numerical flux Obtaining intermediate fluxes by flow vector splitting calculations ;

[0040] Based on the mixing parameter θ, the intermediate flux and numerical flux , the calculated numerical flux is .

[0041] Among them, the intermediate flux The calculation method is:

[0042] ;

[0043] Numerical flux of the positive The calculation method is:

[0044] ;

[0045] in, L i is the eigenvector matrix corresponding to the spatial discrete node i, for L i The inverse matrix of L i+1 is the eigenvector matrix corresponding to the spatial discrete node i+1, for L i+1 The inverse matrix of .

[0046] One or more technical solutions provided by the present invention have at least the following technical effects or advantages:

[0047] The restriction condition proposed by the present invention has less restrictions on the pseudo time step and does not need to be used in every physical time step, so it has high computational efficiency;

[0048] Compared with existing methods, this method can maintain the original high-order physical flux as much as possible, so it has higher calculation accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of the present invention, and do not constitute a limitation on the embodiments of the present invention;

[0050] Figure 1 It is a flow chart of a method for preserving the spatial-temporal correlation scalar based on dual time stepping;

[0051] Figure 2 This is a schematic diagram of density contours of Δt=0.001 of the present invention;

[0052] Figure 3 It is a schematic diagram of density contour lines of Δt=0.0001 of the present invention;

[0053] Figure 4 This is a schematic diagram comparing the density profile of this method with that of references and numerical accurate solutions. DETAILED DESCRIPTION

[0054] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0055] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those within the scope of this description. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0056] Embodiment 1;

[0057] Please refer to Figure 1 , Figure 1 The present invention provides a method for preserving space-time correlation scalars based on dual time-step advancement, and the method comprises:

[0058] Step 1: Start the fluid mechanics numerical simulation, perform numerical discretization, and calculate the physical flux at the unit interface i-1. , the physical flux at the unit interface i and the physical flux at the cell interface i+1 Reconstruct the numerical flux at the unit interface i+1 / 2 ,in, i is the ith spatial discrete node, m For the m Take a walk at a time;

[0059] Step 2: Set the numerical flux at the cell interface i+1 / 2 The conserved variables corresponding to the spatial discrete nodes i and i+1 are and Project upward and calculate the diagonal matrices corresponding to the spatial discrete nodes i and i+1 respectively and ,Will Split to obtain the positive eigenvalue diagonal matrix corresponding to the spatial discrete node i and negative eigenvalue diagonal matrix ,Will Split to obtain the positive eigenvalue diagonal matrix corresponding to the spatial discrete node i+1 and negative eigenvalue diagonal matrix ;

[0060] Step 3: Construct the conditions for scalar preservation;

[0061] Step 4: Obtain the pseudo time step Δτ, based on the pseudo time step Δτ, and , determine whether the scalar preservation condition is met;

[0062] If the scalar preservation condition is met, then based on the numerical flux And the corresponding parameters are numerically iterated;

[0063] If the scalar preservation condition is not met, then based on the pseudo time step Δτ, and , solve to obtain the mixing parameter θ, based on the mixing parameter θ and the numerical flux , and obtain the numerical flux of the positive ; then based on the numerical flux of the positive And the corresponding parameters are numerically iterated.

[0064] In the embodiments of the present invention, please refer to Positivity-preserving high order finite difference WENO schemes for compressible Eulerequations. Journal of Computational Physics 2013(242):169-180. for the numerical simulation and numerical iterative calculation parts, and the present invention will not repeat the corresponding description.

[0065] In the numerical simulation of fluid mechanics, the conserved variables Q and convection flux F in the computational space are the corresponding physical quantities (density, pressure, etc.) in the physical space after coordinate transformation, and t represents the physical time. For any position i in the numerical computational space, as long as the restriction is imposed on each numerical flux , we can achieve a positive scalar.

[0066] In order to significantly increase the physical time step, multicomponent flows (usually accompanied by chemical reactions with different characteristic time scales) tend to introduce pseudo-time iterations and use a dual time stepping method to solve the equations.

[0067] Based on the positive coefficient principle, we derive the sufficient conditions for scalar preservation:

[0068] ; (1)

[0069] Δτ and Δt are the pseudo time step and physical time step respectively, and is the numerical flux The diagonal elements of the diagonal matrix are a x From the maximum characteristic wave velocity τ x and τ y The calculation method is:

[0070] ; (2)

[0071] For a given physical time step Δt and numerical flux , as long as we set Δτ to a sufficiently small value, the above positive condition can be met. However, in the actual numerical calculation process, it is difficult to obtain convergence results by setting such a small pseudo time step. Therefore, for a given physical time step Δt, we constrain the pseudo time step Δτ, specifically:

[0072] ; (3)

[0073] If CFL≤0.5, using the Steger-Warming (flow vector splitting) flux can make equation (1) always hold.

[0074] When the pseudo time step Δτ satisfies the constraints of the above formula, the calculation and If they cannot satisfy the restriction relation of Eq. (1), we will use the numerical flux Further Steger-Warming flux (i.e. intermediate flux) to obtain the correct value , the specific method is:

[0075] ; (4)

[0076] ; (5)

[0077] in, and represents the diagonal matrix consisting of positive, negative eigenvalues ​​or zeros obtained by Steger-Warming splitting, Q i , Q i+1 is the conserved variable at nodes i, i+1 during the discretization of the equation. In addition, the mixing parameter θ should satisfy:

[0078] ; (6)

[0079] The specific implementation steps of the present invention are:

[0080] Step 1): Start the fluid mechanics numerical simulation, use a high-order numerical method (such as 5th-order WENOZ) for numerical discretization, and calculate the numerical flux of the unit interface ;

[0081] Step 2): Convert the numerical flux Perform feature projection to obtain conserved variables and , and then calculated by equations (7) to (9) to obtain and , the specific method is:

[0082] ; (7)

[0083] ; (8)

[0084] ; (9)

[0085] Step 3): Use the pseudo time step Δτ in equation (3) (to avoid θ=1 and thus switch to Steger-Warming flux, CFL=0.25 is recommended here) to determine whether equation (1) holds; if it does, the positive flux can be directly calculated numerically; if it does not hold, solve the mixing parameter θ through equation (6) and calculate the positive value through equation (5) .

[0086] The following is an explanation of this method with a specific example:

[0087] This case is a two-dimensional Sedov detonation wave problem, in which the calculation is seriously unstable due to the vacuum or near-vacuum conditions, causing the calculated density and pressure to become negative. This case is often used in numerical calculations to verify the positivity preservation ability of a particular solution. In this problem, the computational domain ranges from [0,0]×[1.1,1.1], where the high pressure region occupies Figure 2 and Figure 3 The computational unit in the lower left corner.

[0088] exist Figure 2 and Figure 3 The numerical density contours for t = 1 are given in , showing that the proposed implicit dual time-stepping method that maintains positivity does provide smooth density contours. Figure 4In the simulation, we found that the simulated density profile is in good agreement with the results of the reference (Positivity-preserving method for high-order conservative schemes solving compressible Euler equations. Journal of Computational Physics 2013(242):169-180.), and the result of Δt=0.0001 is more accurate than that of Δt=0.001. The simulation results show that the method proposed in this invention performs well in terms of computational accuracy and efficiency.

[0089] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0090] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for preserving spatiotemporal correlation scalars based on dual time stepping, characterized in that: The method comprises: Step 1: Start the fluid mechanics numerical simulation, perform numerical discretization, and calculate the physical flux at the unit interface i-1. , the physical flux at the unit interface i and the physical flux at the cell interface i+1 Reconstruct the numerical flux at the unit interface i+1 / 2 ,in, i is the ith spatial discrete node, m For the m Take a walk at a time; Step 2: Set the numerical flux at the cell interface i+1 / 2 The conserved variables corresponding to the spatial discrete nodes i and i+1 are and Project upward and calculate the diagonal matrices corresponding to the spatial discrete nodes i and i+1 respectively and ,Will Split to obtain the positive eigenvalue diagonal matrix corresponding to the spatial discrete node i and negative eigenvalue diagonal matrix ,Will Split to obtain the positive eigenvalue diagonal matrix corresponding to the spatial discrete node i+1 and negative eigenvalue diagonal matrix ; Step 3: Construct the conditions for scalar preservation; Step 4: Obtain the pseudo time step Δτ, based on the pseudo time step Δτ, and , determine whether the scalar preservation condition is met; If the scalar preservation condition is met, then based on the numerical flux And the corresponding parameters are numerically iterated; If the scalar preservation condition is not met, then based on the pseudo time step Δτ, and , solve to obtain the mixing parameter θ, based on the mixing parameter θ and the numerical flux , and obtain the numerical flux of the positive ; then based on the numerical flux of the positive And the corresponding parameters are numerically iterated.

2. A method for preserving spatiotemporal correlation scalars based on dual time stepping according to claim 1, characterized in that: The conditions for scalar preservation are: ; Where Δt is the physical time step, a x for x Characteristic coefficient in direction, Δ ξ is the curve coordinate along ξ The distance of advancement in the direction, I is the unit diagonal matrix.

3. A method for preserving spatiotemporal correlation scalars based on dual time stepping according to claim 2, characterized in that: a x The calculation method is: ; Among them, τ x for x The maximum characteristic wave speed in the direction, τ y for y The maximum characteristic wave speed in the direction; τ x The calculation method is: ; τ y The calculation method is: ; in, u 1 and u 2 are x and y The velocity component in the direction, c is the speed of sound; ξ x , ξ y , η x and η y are coordinate transformation coefficients, and is the transformation coefficient, , ; Dξ and D η is the node spacing in the computational coordinate system.

4. The method for preserving spatiotemporal correlation scalars based on dual time stepping according to claim 1, characterized in that: The pseudo time step Δτ is calculated as: ; Among them, τ x for x The maximum characteristic wave speed in the direction, τ y for y The maximum characteristic wave velocity in the direction, Δt is the physical time step, and CFL is the CFL number.

5. The method for preserving spatiotemporal correlation scalars based on dual time stepping according to claim 1, characterized in that: The mixing parameter θ is calculated as: ; Where Δt is the physical time step, The physical flux Projection on The diagonal matrix after For The diagonal matrix composed of positive eigenvalues ​​obtained after splitting, For The diagonal matrix composed of negative eigenvalues ​​obtained after splitting, The physical flux Projection on The diagonal matrix after .

6. The method for preserving spatiotemporal correlation scalars based on dual time stepping according to claim 1, characterized in that: and The calculation method is: ; in, L i is the eigenvector matrix corresponding to the spatial discrete node i, for L i The inverse matrix of L i+1 is the eigenvector matrix corresponding to the spatial discrete node i+1, for L i+1 The inverse matrix of and The calculation method is: ; ; The square brackets [ ] contain matrices, and the subscript r represents the rth element in the matrix, It represents No. r elements.

7. The method for preserving spatiotemporal correlation scalars based on dual time stepping according to claim 1, characterized in that: Numerical flux of the positive The calculation method is: The numerical flux Obtaining intermediate fluxes by flow vector splitting calculations ; Based on the mixing parameter θ, the intermediate flux and numerical flux , the calculated numerical flux is .

8. A method for preserving spatiotemporal correlation scalars based on dual time stepping according to claim 7, characterized in that: Intermediate Flux The calculation method is: ; Numerical flux of the positive The calculation method is: ; in, L i is the eigenvector matrix corresponding to the spatial discrete node i, for L i The inverse matrix of L i+1 is the eigenvector matrix corresponding to the spatial discrete node i+1, for L i+1 The inverse matrix of .