Flow guide method for optimizing flue gas flow field treatment

Abstract

The invention discloses a flow guide method for optimizing flue gas flow field treatment, which comprises the following steps of: modeling a flue gas flow guide position by using geometric model establishment, grid division, CFD (Computational Fluid Dynamics) physical model selection, introduction of a physical model with non-Newtonian fluid characteristics, use of a CFD solver, result analysis and optimization adjustment; and different attributes of the non-Newtonian fluid and the Newtonian fluid are simulated respectively, so that the setting accuracy and the setting effect of the guide plate are further ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of waste incineration flue gas diversion treatment, and specifically to a diversion method for optimizing the governance of the flue gas flow field. Background Technique

[0002] In the "arch" part of the boiler flue, the flue gas corridor flow field is prone to unevenness, resulting in uneven flue gas flow through the front and rear sections. The flue gas in the rear section carries a relatively large amount of dust. The dust-containing gas flow has a relatively high velocity at the turning point, causing abrasion to the flue at the turning point and the heating surface tube screen at the lower part of the chimney shaft, affecting the service life of the equipment. During operation, the heating surface at the lower part of the chimney shaft is affected by the uneven flow field, and the ash blockage in the rear section is greater than that in the front section. When the ash blockage is excessive, the system differential pressure increases, the fan power consumption increases, the unit efficiency decreases, and there are great potential safety hazards in operation. The conventional treatment method is to use high-pressure water flushing for the ash-blocked area during unit maintenance, and regularly replace the thinned tube screens and flues and other equipment.

[0003] Existing methods such as the "Method for Optimizing the Flow Field Design of the SCR Inlet Hood Flue" with the patent number ZL202011597148.X already have a method of using CFD simulation to design the diversion. However, in the arch area, the flue gas corridor flow field is prone to unevenness, resulting in uneven flue gas flow through the front and rear sections. The flue gas in the rear section carries a relatively large amount of dust. The dust-containing gas flow has a relatively high velocity at the turning point, which is different from the flue gas situation in the SCR inlet hood flue flow field. In addition, when the flue gas contains a large amount of large particulate matter or droplets, the flue gas will also switch from a Newtonian fluid to a non-Newtonian fluid. Therefore, in the face of the constantly changing fluid in the arch, this method cannot be perfectly applied and there are certain error properties. Summary of the Invention

[0004] The present invention provides a diversion method for optimizing the governance of the flue gas flow field. Compared with the existing waste incineration power plant wastewater treatment method, the present technical solution mainly solves the following technical problems: 1. In the arch area, the flue gas corridor flow field is prone to unevenness, resulting in uneven flue gas flow through the front and rear sections. The flue gas in the rear section carries a relatively large amount of dust. The dust-containing gas flow has a relatively high velocity at the turning point, which is different from the flue gas situation in the SCR inlet hood flue flow field. 2. When the flue gas contains a large amount of large particulate matter or droplets, the flue gas will also switch from a Newtonian fluid to a non-Newtonian fluid. Therefore, in the face of the constantly changing fluid in the arch, the existing methods cannot be applied and there are certain error properties.

[0005] To solve the above technical problems, the technical solution adopted by the present invention is: A diversion method for optimizing the governance of the flue gas flow field, comprising the following steps: S1. Geometric model establishment: Accurately model the overall structure including the deflector and the "arch" flue through a three-dimensional model; S2. Mesh Generation: Import the established geometric model into the mesh generation software for mesh generation. For the flow field area inside the flue, a combination of structured and unstructured meshes is adopted, and local mesh refinement is carried out in key areas such as near the deflector and the wall boundary layer to accurately capture the flow details and gradient changes of the fluid in these areas; S3. Selection of CFD Physical Model: Select a suitable CFD physical model. For the flue gas flow inside this flue, considering its incompressible and turbulent characteristics, the standard k-ε turbulence model is adopted. By solving the governing equations, the values of the turbulent kinetic energy k and the turbulent kinetic energy dissipation rate ε at different positions in the entire flue flow field are gradually iteratively calculated, and then the turbulence-related parameters are determined, so as to accurately describe the turbulent flow state inside the flue under the given boundary conditions; S4. Introduction of the Physical Model with Non-Newtonian Fluid Characteristics: Considering that the flue gas switches between Newtonian and non-Newtonian fluids due to the substances it contains, a physical model with non-Newtonian fluid characteristics is introduced into the CFD physical model for correction, including the power-law model. The expression of the power-law model is: ; where, is the shear stress, is the shear rate, is the consistency coefficient, is the rheological index; S5. Use of CFD Solver: Iteratively solve the set-up model; S6. Result Analysis and Optimization Adjustment: Post-process and analyze the calculated results, adjust the deflector parameters according to the results, and repeat S2~S6 until an optimized design scheme that meets the requirements is obtained.

[0006] Preferably, the S1 includes the following steps: S101. Intelligent Modeling Based on Parametric Design: In the 3D modeling software, make full use of the parametric design function, define the key dimensions of the flue and the deflector as variable parameters, and form an intelligent model by establishing the correlation relationship between the parameters; S102. Combination of Reverse Engineering and Forward Design: If there is an actual flue or a physical object of a similar structure, the point cloud data of the physical object can be first obtained by using 3D scanning technology, then imported into the reverse engineering software for processing to generate an initial geometric model, and then imported into the professional modeling software for forward design optimization; S103. Consideration of Multi-Physical Field Coupling Modeling: Break through the traditional modeling idea of only considering fluid mechanics, consider multi-physical field coupling factors in the geometric model establishment stage, realize the preliminary construction of the multi-physical field model, and make the simulation results closer to the actual complex working conditions.

[0007] Preferably, S2 includes the following steps: Set up a boundary layer grid near the surface of the deflector. The height of the first layer is set to 0.1 mm, the total number of layers is 10, and the growth rate is 1.2 to ensure that the y+ value near the wall is within a reasonable range. The overall number of grids is reasonably adjusted according to the complexity of the model and computing resources, ranging from millions to tens of millions of cells, so as to improve the computing efficiency as much as possible on the premise of ensuring the computing accuracy.

[0008] Preferably, in the power-law model in S4, when , the power-law model degenerates into a Newtonian fluid model; when , the fluid behaves as a pseudoplastic fluid, that is, as the shear rate increases, the viscosity decreases; when , the fluid behaves as a dilatant fluid, and as the shear rate increases, the viscosity increases.

[0009] More preferably, in the power-law model in S4, the consistency coefficient takes a value of 1.0, the rheological index takes any one of the values 0.5, 1.0, and 1.5, and the shear rate takes 100 numbers with equal numerical intervals distributed between 0 and 10.

[0010] Preferably, the physical model of the non-Newtonian fluid characteristics in S4 also includes the Bingham plastic model, which is applicable to describe fluids with a yield stress. Its mathematical expression is: ; where is the yield stress, and only when the shear stress exceeds the yield stress will the fluid flow, is the plastic viscosity, is the shear stress, is the shear rate.

[0011] More preferably, in the Bingham plastic model in S4, the yield stress takes a value of 1.0, representing the critical shear stress, and the plastic viscosity takes a value of 0.5.

[0012] Preferably, S4 also includes the following steps: S401. After determining the non-Newtonian fluid parameters, incorporate them into the CFD simulation. The CFD simulation solves the control equations of fluid flow through numerical calculations. For non-Newtonian fluids, the viscosity term in the control equations needs to be corrected to consider the characteristics of non-Newtonian fluids. In this way, the flow behavior of flue gas in the flue can be predicted more accurately; S402. Physical model correction based on data-driven approach: Correct the physical model using experimental data or on-site monitoring data. Set up multiple monitoring points in the actual flue gas system to collect physical quantity data in real-time. Compare and analyze these measured data with the CFD simulation results. Establish a data-driven correction model using machine learning algorithms. Adjust the empirical coefficients in the standard k-ε turbulence model in real-time according to the measured data to make the simulation results closer to the actual situation and improve the accuracy and reliability of the model.

[0013] Preferably, the S5 includes the following steps: Set the residual convergence criterion of the energy equation to 1×10 -6 , and the residual convergence criteria of the continuity equation and the momentum equation to 1×10 -4 . At the same time, monitor the changes in physical quantities at key positions to ensure the stability and accuracy of the calculation results. During the calculation process, reasonably adjust the number of iteration steps and the time step according to the computing resources and convergence situation. Set the initial time step to 0.001 s, and increase the time step as the calculation progresses to accelerate the convergence speed, but ensure the stability of the calculation.

[0014] A flow guiding device for optimizing the flue gas flow field. Using the above-mentioned flow guiding method for the flow guiding device arranged for the arch-shaped chimney, it includes a flow guiding plate. The inner arc guiding angle of the flow guiding plate is 60°, and the outer arc guiding angle is 20°. The flow guiding plate is fixed to the flue gas truss pipe by welding triangular plates. The triangular plates are welded to the leeward side of the pipe, and the flue gas direction is arranged at both ends of the flow guiding plate.

[0015] Advantages of the present invention: (1) The flue gas flow guiding arrangement scheme obtained by this method will reduce the high-speed area in the arch-shaped flue, reducing the wear of dust on the flue, support pipes, and tube screens; after the flow field becomes uniform, the deposition of fly ash in the high-speed area decreases, and the flow resistance drops; the unit operates stably economically, and the efficiency is guaranteed.

[0016] (2) In response to the switching of flue gas between Newtonian fluid and non-Newtonian liquid properties, a physical model with non-Newtonian fluid characteristics is introduced, optimizing the original CFD physical model, further enhancing the applicability and accuracy, and enabling the generated scheme to be better applicable to the corresponding flue. Description of the drawings

[0017] Figure 1 It is a model comparison diagram of the flow field before the implementation of the flow guiding (left) and after the implementation of the flow guiding (right) of the present invention. Detailed implementation manners

[0018] As follows, the embodiments will be further described with reference to the drawings.

[0019] As a preferred Embodiment 1, a diversion method for optimizing the flue gas flow field governance includes the following steps: S1. Geometric model establishment: Accurately model the overall structure including the deflector and the "arch bridge" flue through a three-dimensional model; S2. Mesh generation: Import the established geometric model into the mesh generation software for mesh generation. For the flow field area inside the flue, a combination of structured or unstructured meshes is adopted, and local mesh refinement is carried out in key areas such as near the deflector and the wall boundary layer to accurately capture the flow details and gradient changes of the fluid in these areas; S3. CFD physical model selection: Select a suitable CFD physical model. For the flue gas flow inside this flue, considering its incompressible and turbulent characteristics, the standard k-ε turbulence model is adopted. By solving the governing equations, the values of the turbulent kinetic energy k and the turbulent kinetic energy dissipation rate ε at different positions in the entire flue flow field are gradually iteratively calculated, and then the turbulence-related parameters are determined, so as to accurately describe the turbulent flow state inside the flue under the given boundary conditions; S4. Introduction of the physical model of non-Newtonian fluid characteristics: Considering that the flue gas switches between Newtonian fluid and non-Newtonian fluid due to the substances it contains, a physical model of non-Newtonian fluid characteristics is introduced into the CFD physical model for correction, including the power-law model. The expression of the power-law model is: ; where is the shear stress, is the shear rate, is the consistency coefficient, is the rheological index; S5. Use of the CFD solver: Iteratively solve the set model; S6. Result analysis and optimization adjustment: Post-process and analyze the calculated results, adjust the deflector parameters according to the results, and repeat S2~S6 until an optimized design scheme that meets the requirements is obtained.

[0020] As a preferred Embodiment 2, the S1 includes the following steps: 1. Intelligent modeling based on parametric design: In the three-dimensional modeling software, make full use of the parametric design function to define the key dimensions of the flue and the deflector, such as length, radius, angle, etc., as variable parameters. By establishing the correlation relationship between the parameters, an intelligent model is formed. The user only needs to input different working condition parameters, and the model can be automatically updated without redrawing every detail.

[0021] 2. Combination of Reverse Engineering and Forward Design: If there is a physical object of an actual flue or a similar structure, the point cloud data of the physical object can be obtained first using 3D scanning technology, and then imported into reverse engineering software such as Geomagic for processing to generate an initial geometric model. Then it is imported into professional modeling software for forward design optimization, such as smoothing the curved surface of the "arch bridge" flue and precisely adjusting the position and shape of the deflector.

[0022] 3. Consideration of Multi-Physical Field Coupling Modeling: Break through the traditional modeling idea that only considers fluid mechanics, and consider multi-physical field coupling factors at the stage of establishing the geometric model. For example, if there is a temperature gradient in the flue, which affects the density and viscosity of the flue gas, temperature-related parameter regions can be set in the model to lay the foundation for subsequent thermal-fluid-structure interaction analysis. By creating a virtual temperature field region in the modeling software and combining it with the fluid flow region, a preliminary multi-physical field model is constructed, making the simulation results closer to the actual complex working conditions.

[0023] 4. Modeling Optimization Based on Artificial Intelligence Assistance: Use artificial intelligence algorithms to assist in modeling optimization. After completing the preliminary geometric model, input the geometric parameters of the model and some expected performance indicators (such as pressure loss, flow uniformity, etc.) into the trained neural network model. Through artificial intelligence algorithms, predict the impact of different geometric parameter adjustments on performance indicators, thereby guiding the optimization of the shape and position of the flue and deflector in the modeling software. For example, through neural network prediction, it is found that increasing the inclination angle of the deflector by 5° can increase the flow uniformity at the flue outlet by 10%, and then corresponding adjustments are made in the modeling software to achieve efficient modeling optimization based on intelligent algorithms.

[0024] As a preferred embodiment 3, S2 includes the following steps: Import the established geometric model into a mesh generation software (such as ANSYS Meshing, ICEM CFD, etc.) for mesh generation. For the flow field region in the flue, a combination of structured and unstructured meshes is adopted, and local mesh refinement is carried out in key regions such as near the deflector and the wall boundary layer to accurately capture the flow details and gradient changes of the fluid in these regions. For example, set a boundary layer mesh near the surface of the deflector, with the first layer height set to 0.1 mm, the total number of layers to 10, and the growth rate to 1.2, to ensure that the y+ value near the wall is within a reasonable range. The overall number of meshes is reasonably adjusted according to the complexity of the model and computing resources, generally between several million and tens of millions of elements, to improve the computing efficiency as much as possible while ensuring the computing accuracy.

[0025] 1. Adaptive grid dynamic adjustment: Traditional grid division is often completed once before the simulation begins, but during the flow of smoke in the flue, key areas may change at different times. Develop an adaptive grid dynamic adjustment algorithm to monitor flow field parameters (such as velocity gradient, pressure gradient, etc.) in real time during the simulation. When the parameter changes in certain areas exceed a certain threshold, the grid of the area is automatically encrypted; when the parameter changes tend to be flat, the grid is appropriately coarsened. For example, in the process of vortex formation or disappearance in the flue, the grid density around it is adjusted in real time according to the dynamic changes of the vortex. This can not only accurately capture the dynamic changes of the flow field, but also avoid using too many grids in unnecessary areas, thereby improving computational efficiency.

[0026] 2. Mesh generation strategy based on machine learning: Use machine learning algorithms to optimize mesh generation strategies. Collect a large number of meshing cases under different flue structures and flow field conditions and their corresponding simulation results (such as calculation accuracy, calculation time, etc.) as training data. Train a machine learning model, such as a deep neural network, so that it can predict the optimal meshing scheme based on the input flue geometric characteristics, flow field boundary conditions and other information, including mesh type (structured or unstructured), encrypted area, mesh size, etc. In practical applications, you only need to input the new flue model and boundary conditions into the trained model to quickly obtain a suitable meshing strategy, reducing the time and cost of manual trial and error.

[0027] 3. Multi-scale grid fusion technology: For the flue system, different parts may have different characteristic scales. The scale of the main part of the flue is large, while the scale of structures such as guide plates, local protrusions or depressions is small. Multi-scale grid fusion technology can be used to use relatively coarse grids in large-scale areas to reduce the amount of calculation; use fine grids in small-scale areas to capture local flow details. Through special grid transition methods, smooth connections between grids of different scales are ensured to avoid numerical calculation errors caused by large differences in grid scales. This method can significantly reduce the overall number of grids and improve calculation efficiency while ensuring calculation accuracy.

[0028] As a preferred embodiment 4, the power law model in S4 is When , the power law model degenerates into a Newtonian fluid model; when When , the fluid behaves as a pseudoplastic fluid, that is, as the shear rate increases, the viscosity decreases; when When the shear rate increases, the fluid behaves as an expansive fluid, and the viscosity increases with the increase of shear rate.

[0029] To accurately describe the flow behavior of flue gas in the flue duct, especially under conditions of high particle concentration or special flow conditions, it is necessary to determine the non-Newtonian fluid parameters of the flue gas under the current operating conditions. This can be achieved through experiments or theoretical analysis. Experimental methods usually include measurements with a rotational rheometer, capillary rheometer, etc. Taking the rotational rheometer as an example, by measuring the torque exerted on the fluid at different rotation speeds, the shear stress and shear rate are then calculated, and the corresponding non-Newtonian fluid parameters are fitted based on the experimental data. Theoretical analysis requires combining multidisciplinary knowledge such as fluid mechanics and particle dynamics, considering factors such as the interaction between particles and the fluid, the concentration distribution of particles, and the particle size distribution, to establish a mathematical model to deduce the non-Newtonian fluid parameters.

[0030] As a more preferred option, in the power-law model in S4, the consistency coefficient takes a value of 1.0, and the rheological index takes any one of the values 0.5, 1.0, and 1.5, and the shear rate takes 100 numbers distributed at equal numerical intervals between 0 and 10.

[0031] As a preferred embodiment 5, the physical model of the non-Newtonian fluid characteristics in S4 further includes the Bingham plastic model, which is applicable to describing fluids with a yield stress, and its mathematical expression is: ; where is the yield stress, and only when the shear stress exceeds the yield stress will the fluid flow, is the plastic viscosity, is the shear stress, is the shear rate.

[0032] As a more preferred option, in the Bingham plastic model in S4, the yield stress takes a value of 1.0, representing the critical shear stress, and the plastic viscosity takes a value of 0.5.

[0033] Embodiment 4 and Embodiment 5 can be expressed as the following code: python import numpy as np import matplotlib.pyplot as plt # Power-law model def power_law_model(shear_rate, K, n): return K * shear_rate ** n # Bingham plastic model def bingham_plastic_model(shear_rate, tau_y, mu_p): return np.where(shear_rate == 0, 0, tau_y + mu_p * shear_rate) # Parameter settings # Power-law model parameters K = 1.0 n_values = [0.5, 1.0, 1.5] # Bingham plastic model parameters tau_y = 1.0 mu_p = 0.5 # Shear rate range shear_rate = np.linspace(0, 10, 100) # Plot the power-law model plt.figure(figsize=(12, 6)) for n in n_values: shear_stress = power_law_model(shear_rate, K, n) if n == 1: label = f'Newtonian fluid (n={n})' elif n<1: label = f'Pseudoplastic fluid (n={n})' else: label = f'Dilatant fluid (n={n})' plt.plot(shear_rate, shear_stress, label=label) # Plot the Bingham plastic model shear_stress_bingham = bingham_plastic_model(shear_rate, tau_y, mu_p) plt.plot(shear_rate, shear_stress_bingham, label='Bingham plastic model', linestyle='--') plt.xlabel('Shear rate ($\\dot{\\gamma}$)') plt.ylabel('Shear stress ($\\tau$)') plt.title('Comparison between power-law model and Bingham plastic model') plt.legend() plt.grid(True) plt.show() Code explanation: Power-law model function power_law_model: Calculate the corresponding shear stress according to the given shear rate shear_rate, consistency coefficient K, and flow index n.

[0034] Bingham plastic model function bingham_plastic_model: Calculate the corresponding shear stress according to the given shear rate shear_rate, yield stress tau_y, and plastic viscosity mu_p. Here, the np.where function is used to handle the case where the shear stress is 0 when the shear rate is 0.

[0035] Parameter settings: Set the consistency coefficient K and different flow indices n in the power-law model, as well as the yield stress tau_y and plastic viscosity mu_p in the Bingham plastic model.

[0036] Plot the graph: Use the matplotlib library to plot the relationship curves of shear stress and shear rate for the power-law model and the Bingham plastic model, and add labels and titles to distinguish different models and parameter cases.

[0037] Preferably, step S4 further includes the following steps: S401. After determining the non-Newtonian fluid parameters, incorporate them into the CFD (Computational Fluid Dynamics) simulation. CFD simulation is to numerically solve the governing equations of fluid flow, such as the continuity equation, momentum equation, and energy equation, etc. For non-Newtonian fluids, the viscosity term in the governing equations needs to be corrected to account for the characteristics of non-Newtonian fluids. For example, under the power-law model, the viscosity is substituted into the momentum equation, and then numerical methods such as the finite element method and finite volume method are used for discrete solution to obtain the distributions of physical quantities such as the velocity field, pressure field, and temperature field of the flue gas in the flue. In this way, the flow behavior of the flue gas in the flue can be predicted more accurately.

[0038] S402. Physical model correction based on data-driven approach: The physical model is corrected using experimental data or on-site monitoring data. Multiple monitoring points are set in the actual flue gas system to collect data of physical quantities such as velocity, pressure, and temperature in real time. These measured data are compared and analyzed with the CFD simulation results, and a data-driven correction model is established using machine learning algorithms (such as neural networks, support vector machines, etc.). The empirical coefficients (such as the coefficients of turbulent kinetic energy generation term, dissipation term, etc.) in the standard k-ε turbulence model are adjusted in real time according to the measured data to make the simulation results closer to the actual situation and improve the accuracy and reliability of the model.

[0039] As a preferred embodiment 6, the S5 includes the following steps: Use a CFD solver (such as ANSYS Fluent, CFX, etc.) to perform iterative solution on the set model. Set the residual convergence criterion of the energy equation to 1×10 -6 , and the residual convergence criteria of the continuity equation and momentum equation to 1×10 -4 . At the same time, monitor the changes in physical quantities at key positions (such as the pressure and velocity distributions before and after the deflector) to ensure the stability and accuracy of the calculation results. During the calculation process, reasonably adjust the number of iteration steps and time step size according to the computing resources and convergence situation. Generally, the initial time step size is set to 0.001 s, and the time step size can be appropriately increased as the calculation progresses to accelerate the convergence speed, but the stability of the calculation should be ensured.

[0040] As Figure 1 shown, as a preferred embodiment 7, a deflector device for optimizing the flue gas flow field governance, using the above-mentioned method for optimizing the flue gas flow field governance for the deflector device arranged in the arch-shaped flue shaft, includes a deflector plate. The inner arc deflector angle of the deflector plate is 60°, and the outer arc deflector angle is 20°. The deflector plate is fixed to the flue duct truss pipe by welding triangular plates. The triangular plates are welded to the leeward side of the pipe, and the flue gas direction is arranged at both ends of the deflector plate.

[0041] The ash accumulation reduction structure of this patent application is mainly the heating surface at the lower section of the "arch-shaped" flue shaft. The upper opening size of the cross-section of the arch side is 1000 mm, and the lower opening size is 2670 mm. Two sections of deflector plates are correspondingly configured, with an inner arc deflector angle of 60° and R2384 mm, and an outer arc deflector angle of 20° and R7700 mm. T5 steel plate Q355B is used. The deflector device is fixed to the flue duct truss pipe by welding triangular plates. The triangular plates are welded to the leeward side of the pipe, and the side length is not less than 150 mm. The reinforcing flat steel of the deflector plate is evenly arranged in the flue duct along the width direction with a pitch of 2128 mm, and the flue gas direction is arranged at both ends of the deflector plate and can be adjusted appropriately.

[0042] Number of flow deflectors: two sections. This number setting is based on in-depth research and simulation analysis of the flue gas flow characteristics in the lower heating surface of the "arch bridge" flue shaft. Through the reasonable arrangement of the two-section flow deflectors, the flow direction of the flue gas in the flue can be effectively guided, reducing the turbulence and recirculation of the flue gas, thereby reducing the possibility of ash accumulation.

[0043] Inner arc deflection angle: 60°. This angle is designed to enable the flue gas to form a relatively concentrated and stable flow direction on the inner arc side, which helps to carry the dust particles in the flue gas along a predetermined path and avoid the accumulation of dust in the flue. At the same time, combined with the design of the inner arc radius R2384mm, the flow of the flue gas in the inner arc area can be made smoother, reducing dust deposition caused by factors such as sharp turns.

[0044] Outer arc deflection angle: 20°. The relatively small outer arc deflection angle, combined with the outer arc radius of R7700mm, mainly takes into account the flow characteristics of the flue gas on the outer arc side and its synergistic effect with the flue gas on the inner arc side. Such a design can enable the flue gas to form a relatively uniform flow field distribution as a whole when passing through the flow deflectors, improving the flue gas transportation efficiency and further reducing the generation of ash accumulation.

[0045] Position relationship with the arch bridge: Configured on the lower heating surface of the "arch bridge" flue shaft, corresponding to the upper opening size of 1000mm and the lower opening size of 2670mm on the cross-section of the arch bridge side. The precise position correspondence is determined according to the unique structure of the arch bridge flue and the flow law of the flue gas in this area. By matching the flow deflectors with the specific dimensions of the arch bridge, the flue space can be maximally utilized, the flue gas diversion path can be optimized, and the flow deflectors can better adapt to the flow field changes in the arch bridge flue, thereby achieving the desired high-efficiency effect.

[0046] Compared with the flow guiding components in the prior art, the flow deflectors in this patent have unique designs in terms of angle, radius, fixing method, and position relationship with the arch bridge, and can better adapt to the specific flue structure and flue gas flow requirements, improving the operation efficiency and stability of the equipment.

[0047] As a preferred embodiment 8, this method can also select the CFG simulation specific model, and the specific steps are as follows: I. Computational domain model The computational domain includes the entire flue space and the inlet and outlet extension sections. The length of the inlet and outlet extension sections is generally taken as 3 - 5 times the flue diameter to ensure that the influence of the inlet and outlet boundary conditions on the internal flow field can fully develop. The shape and size of the computational domain are determined according to the actual installation position of the flue and the surrounding environment to ensure that the simulated flow field can reflect the actual flow situation.

[0048] II. Flow deflector model The deflector is accurately constructed in the model with its actual shape and dimensions, including parameters such as the curvature radii of the inner arc and the outer arc, the deflection angle, the length, and the thickness. For example, the inner arc deflection angle is 60°, R is 2384 mm, the outer arc deflection angle is 20°, R is 7700 mm, and the thickness of the deflector is t5. The surface of the deflector is set with a wall boundary condition to interact with the fluid in the flue gas duct and simulate its effect on deflecting the flue gas flow.

Claims

1. A flow guiding method for optimizing the treatment of flue gas flow field, characterized in that, It includes the following steps: S1. Geometric model establishment: Accurately model the overall structure including the deflector and the "arch bridge" flue through a 3D model; S2. Mesh generation: Import the established geometric model into the mesh generation software for mesh generation. For the flow field area inside the flue, a combination of structured or unstructured meshes is adopted, and local mesh refinement is performed in key areas such as near the deflector and the wall boundary layer to accurately capture the flow details and gradient changes of the fluid in these areas; S3. Selection of CFD physical model: Select a suitable CFD physical model. For the flue gas flow inside this flue, considering its incompressible and turbulent characteristics, the standard k-ε turbulence model is adopted. By solving the governing equations, the values of the turbulent kinetic energy k and the turbulent kinetic energy dissipation rate ε at different positions in the entire flue flow field are calculated step by step through iterative calculations, and then the turbulence-related parameters are determined, so as to accurately describe the turbulent flow state inside the flue under the given boundary conditions; S4. Introduction of the physical model with non-Newtonian fluid characteristics: Considering that the flue gas switches between Newtonian and non-Newtonian fluids due to the substances it contains, a physical model with non-Newtonian fluid characteristics is introduced into the CFD physical model for correction, including the power-law model. The expression of the power-law model is: ； Among them, is the shear stress, is the shear rate, is the consistency coefficient, is the rheological index; S5. Use of CFD solver: Iteratively solve the set-up model; S6. Result analysis and optimization adjustment: Post-process and analyze the calculated results, adjust the deflector parameters according to the results, and repeat S2 to S6 until an optimized design scheme that meets the requirements is obtained.

2. The diversion method for optimizing the flue gas flow field treatment according to claim 1, characterized in that, The S1 includes the following steps: S101. Intelligent modeling based on parametric design: In the 3D modeling software, make full use of the parametric design function, define the key dimensions of the flue and the deflector as variable parameters, and form an intelligent model by establishing the correlation relationship between the parameters; S102. Combination of reverse engineering and forward design: If there is an actual flue or a physical object of a similar structure, the point cloud data of the physical object can be obtained first using 3D scanning technology, then imported into the reverse engineering software for processing to generate an initial geometric model, and then imported into the professional modeling software for forward design optimization; S103. Consideration of multi-physics field coupling modeling: Break through the traditional modeling idea that only considers fluid mechanics, consider multi-physics field coupling factors in the geometric model establishment stage, and realize the preliminary construction of the multi-physics field model to make the simulation results closer to the actual complex working conditions.

3. The diversion method for optimizing the flue gas flow field governance according to claim 1, characterized in that The S2 includes the following steps: Set up boundary layer meshes near the surface of the deflector, with the first layer height set to 0.1 mm, the total number of layers to 10 layers, and the growth rate to 1.2 to ensure that the y+ value near the wall is within a reasonable range. The overall number of meshes is reasonably adjusted according to the complexity of the model and the computing resources, between millions and tens of millions of elements, so as to improve the computing efficiency as much as possible while ensuring the computing accuracy.

4. A flow guiding method for optimizing the flue gas flow field treatment according to claim 1, characterized in that The power-law model in S4 degenerates into the Newtonian fluid model when ; when , the fluid behaves as a pseudoplastic fluid, i.e., the viscosity decreases as the shear rate increases; when , the fluid behaves as a dilatant fluid, and the viscosity increases as the shear rate increases.

5. A flow guiding method for optimizing the flue gas flow field treatment according to claim 4, characterized in that The consistency coefficient in the power-law model in S4 takes a value of 1.0, and the flow index takes any one of the values 0.5, 1.0, and 1.5, and the shear rate takes 100 numbers with equal numerical intervals distributed between 0 and 10.

6. The guiding method for optimizing the flue gas flow field treatment according to claim 1, characterized in that The physical model with non-Newtonian fluid characteristics in S4 also includes the Bingham plastic model, which is applicable to describe fluids with yield stress. Its mathematical expression is: ； wherein is the yield stress. Only when the shear stress exceeds the yield stress will the fluid flow. is the plastic viscosity. is the shear stress. is the shear rate.

7. A flow guiding method for optimizing the flue gas flow field governance according to claim 6, characterized in that, The yield stress in the Bingham plastic model of S4 takes a value of 1.0, representing the critical shear stress, and the plastic viscosity takes a value of 0.

5.

8. A flow guiding method for optimizing the flue gas flow field governance according to claim 1, characterized in that, The S4 also includes the following steps: S401. After determining the non-Newtonian fluid parameters, incorporate them into the CFD simulation. The CFD simulation solves the control equations of fluid flow through numerical calculations. For non-Newtonian fluids, the viscosity term in the control equations needs to be corrected to account for the characteristics of non-Newtonian fluids. In this way, the flow behavior of flue gas in the flue can be predicted more accurately. S402. Data-driven physical model correction: Use experimental data or on-site monitoring data to correct the physical model. Set multiple monitoring points in the actual flue system to collect physical quantity data in real time. Compare and analyze these measured data with the CFD simulation results. Use machine learning algorithms to establish a data-driven correction model, and adjust the empirical coefficients in the standard k-ε turbulence model in real time according to the measured data to make the simulation results closer to the actual situation and improve the accuracy and reliability of the model.

9. The diversion method for optimizing the flue gas flow field governance according to claim 1, characterized in that, The S5 includes the following steps: Set the residual convergence criterion of the energy equation to 1×10 -6 , and the residual convergence criteria of the continuity equation and the momentum equation to 1×10 -4 . At the same time, monitor the changes in physical quantities at key positions to ensure the stability and accuracy of the calculation results. During the calculation process, reasonably adjust the number of iteration steps and the time step according to the computing resources and convergence conditions. The initial time step is set to 0.001 s, and the time step is increased as the calculation progresses to accelerate the convergence speed, but the stability of the calculation must be ensured.

10. A flow guiding device for optimizing the treatment of flue gas flow field, characterized in that, Use a flow guiding method for optimizing the flue gas flow field governance according to any one of claims 1 to 9 for the flow guiding device arranged for the arch bridge flue well, including a flow guiding plate. The inner arc flow guiding angle of the flow guiding plate is 60°, and the outer arc flow guiding angle is 20°. The flow guiding plate is fixed to the flue truss pipe by welding triangular plates. The triangular plates are welded to the leeward side of the pipe, and the flue gas direction is arranged at both ends of the flow guiding plate.

Images