Numerical Simulation of the Flow Field in a Tubular Thermal Cracking Reactor for Water Vapor and Difluoromonochloromethane

Abstract

Tetrafluoroethylene (TFE), as a key basic chemical raw material, has an irreplaceable position in strategic emerging industries involving high-end materials, electronics, chemicals, and pharmaceuticals. Currently, TFE is industrially produced via the vapor cracking of difluoromonochloromethane (R22). However, there is a gap between China and the developed countries in the high-end tetrafluoroethylene monomer, the purity of tetrafluoroethylene monomer is difficult to reach the high purity requirement of 99.999%, and the content of the key impurities that determine the nature of the functional materials is high, which leads to a series of problems of instability in the performance of the high-end and special products and high media loss. To enhance the purity of TFE monomers produced by the pyrolysis reactor of R22 and water vapor, the fluid dynamics simulations of the reactor model were conducted using Ansys Fluent. The reactor model was initially constructed using Space Claim, followed by mesh generation with Fluent Meshing and other relevant configurations. Both cold-state and thermal-state simulations were performed. The cold-state simulation analyzed the effects of temperature, flow velocity, and turbulence models on the turbulent gas flow and mixing processes within the reactor model. The thermal-state simulation examined the impacts of reaction process variations on internal temperature, turbulence, component distribution, and outlet component concentrations during the actual reaction process. Finally, the inlet flow rate and structure of the reactor were optimized. The results indicated that the optimal inlet flow rates for R22 and water vapor were 0.2–0.3 kg/s and 0.4–0.5 kg/s, respectively. In practical production, the internal fluid mixing achieved an optimal value after modifying the inlet structure to a T shape. This study provides new insights into the pyrolysis reaction and lays the foundation for further improving the purity of TFE monomers.

1. Introduction

TFE (chemical formula: C₂F₄) is a colorless, odorless gas. Due to its unique chemical and physical properties, it is widely used in chemical industry, medicine, and material science [1,2,3]. Tetrafluoroethylene as the main raw material for the production of PTFE, its purity directly affects the performance and application range of the final product. The purity of Tetrafluoroethylene monomer plays a decisive role in the quality of PTFE resin, and only when the quality of monomer is stable, the quality of resin will not fluctuate greatly. At the same time, in order to meet the requirements of high-end customers in pursuit of high strength and low shrinkage, we should pay special attention to the purification and purification of monomer, and remove the impurities in monomer as far as possible. However, there is still a gap between China’s high-end tetrafluoroethylene monomer and developed countries. Domestic products are concentrated in the low-end range, and the 99.999% purity requirement for monomers is difficult to meet, while the high-end products, especially those used in extreme environments, are involved in the key areas of national strategy and are subjected to strict technological embargoes and the monopolization of the product market by DuPont of the U.S.A. and Daikin of Japan. This pertains to the critical national priority of urgently addressing the “chokepoint” issues of basic raw materials that require immediate resolution [4].

Currently, the primary method for industrial production of TFE involves the dilution cracking of R22 using water vapor [5]. This reaction typically takes place in a tubular cracking reactor. The entire cracking process is exothermic, meaning it releases heat. During the reaction, difluorocarbene groups are generated, which are highly reactive and make it difficult to control the reaction conditions. Thus, a significant amount of by-products are often produced within the reactor [6,7]. This increase in by-products complicates the purification of TFE and significantly hampers advancements in the production of high-purity TFE.

The aim of the study of the thermal cracking reactor designed for the vaporization of R22 is to increase the purity of the main product, TFE, while minimizing the formation of by-products. This aim is pursued by constantly adjusting the reaction conditions; however, this approach presents several challenges in practice [8]. The cracking reaction takes place at high temperatures, typically exceeding 700 °C and reaching 750 °C to 790 °C, with a surface residence time of only about 0.04 s. In view of the high and difficult-to-control reaction temperatures, this study aims to propose a new method to simulate the thermal cracking reactor of R22 in water vapor using a fluid dynamics software, to simulate the process and optimize the conditions using the Ansys Fluent 2022 Fluid Dynamics software, and to provide a theoretical basis for the actual production of the reactor in order to improve the purity of TFE monomers.

2. Models and Methods

2.1. Mathematical Model

In this study, we assume that the cracking reactor operates in a steady state with stable parameters. The mass flow rates of R22 and water vapor remain constant over time. Additionally, the reactor is in a fully turbulent state, and the reaction occurring inside the reactor is rapid, quickly reaching an equilibrium [9]. This study focuses on the numerical simulation of a two-dimensional, steady-state, incompressible, turbulent reaction process. We perform the numerical simulation of the cracking reactor based on the equations governing fluid flow using Ansys Fluent.

In this study, we explore the thermal cracking process of R22 in the presence of water vapor. This process is coupled to a component transfer model that incorporates volumetric reactions. The component transfer behavior includes the processes of diffusion, convection, and reaction. Therefore, the modeling and calculations are grounded in the principles of mass conservation and component conservation. In the mathematical model established by the software, each component is assigned a mass fraction to describe its proportion within the fluid [10].

2.1.1. Continuity Equation

The continuity equation is a specific expression of the law of conservation of mass in fluid flow systems. All fluid flow problems must satisfy this equation, which states that the increase in mass in a fluid microcosm per unit of time is equal to the net mass flowing into the microcosm during the same time interval. In this model, the flow of the fluid remains constant over time, and the fluid is assumed to be a gas. Thus, the density ($\rho$) of the fluid does not change throughout the process. From the law of conservation of mass, we can derive the mass conservation equation [11].

$${\displaystyle \frac{\partial \mathrm{\rho }}{\partial \mathrm{t}}}+{\displaystyle \frac{\partial }{\partial {\mathrm{x}}_{\mathrm{i}}}}\left(\mathrm{\rho }{\mathrm{v}}_{\mathrm{i}}\right)=0$$

(1)

where $i$ is the coordinate direction, and ${\mathrm{v}}_{\mathrm{i}}$ is the gas phase velocity.

2.1.2. Momentum Equations

The momentum equation is a fundamental equation that describes the motion of a fluid. It expresses the conservation of momentum over time through any cross-section of the flow. This equation serves as a basic principle that must be upheld by any flow system. It states that the rate of change in fluid momentum over time in any controlled microelement is equal to the sum of external forces acting on that microelement [12].

$${\displaystyle \frac{\partial }{\partial t}}\left(\rho v_i\right)+{\displaystyle \frac{\partial }{\partial x_j}}\left(\rho v_i{v}_j\right)=-{\displaystyle \frac{\partial p}{\partial x_i}}+{\displaystyle \frac{\partial {\tau }_{ij}}{\partial x_j}}+\rho g_i+F_i$$

(2)

where i and j are directions, v_i and v_j are phase velocities, τ_ij is the stress tensor, g_i and F_i are the gravitational volumetric force and the external volumetric force in the i direction, respectively, and F_i contains other model-related source terms.

2.1.3. Energy Equation

The energy equation is a fundamental principle that describes heat transfer in fluids. It expresses the conservation of energy per unit time across any interface and serves as a basic law that must be upheld in any flow system involving heat exchange. This law states that the rate of energy increase within a defined volume, or microcosm, is equal to the net heat flux entering that volume plus the work performed by both volumetric and surface forces.

The energy E of a fluid is usually the sum of the three terms: internal energy i, kinetic energy $K={\displaystyle \frac{1}{2}}\left(u^2+v^2{+w}^2\right),$ and potential energy P. There is a one-point relationship between the internal energy i and the temperature T, i = c_pT, where c_p is the specific heat capacity. This gives the energy conservation equation with temperature T as the variable:

$${\displaystyle \frac{\partial \left(\rho T\right)}{\partial t}}+div\left(\rho uT\right)=div\left({\displaystyle \frac{k}{c_p}}gradT\right)+S_T$$

(3)

where c_p is the specific heat capacity, T is the temperature, k is the heat transfer coefficient of the fluid, and S_T is the heat source within the fluid and the part of the fluid mechanical energy converted to thermal energy due to viscous action [13,14].

2.2. Physical Model

In Ansys Fluent, physical modeling is essential for simulating fluid flow and associated physicochemical processes. The software offers a diverse range of physical models that cover various applications, from incompressible to compressible flow, laminar to turbulent flow, and heat transfer and phase change. It also supports chemical reactions and combustion, multiphase flow, rotating machinery, dynamic or deformable meshes, noise analysis, material processing, fuel cell simulations, etc. In this study, we analyze a thermal cracking reactor model using R22 and water vapor. We select the component transfer model with turbulence–chemical reaction interactions in volumetric reactions, ensuring that the relevant setup is properly configured for our analysis [15,16,17].

2.3. Reaction Model

Difluorocarbene is the primary influencing species in the cracking reaction of R22 and serves as an important active intermediate in organic fluorine chemistry. A carbene is an electrically neutral compound that contains divalent carbon, meaning the carbon is covalently bonded to only two other groups, with two unbonded electrons remaining. In the case of difluorocarbene, there are two key effects at play. First, the strong electronegative influence of the fluorine atom creates a significant electron deficiency at the carbon center. Secondly, the lone pair of electrons on the fluorine atom can fill the empty p-orbitals of difluorocarbene when it adopts a linear configuration, which helps to reduce the electron deficiency at the carbon center. The simultaneous presence of these two factors makes difluorocarbene a moderately electrophilic species [18,19,20]. Difluorocarbene is extremely reactive and exists as a short-lived organic intermediate. In fact, carbenes are generally more unstable than carbocations or radicals. When formed in a reaction mixture, difluorocarbene can quickly react with other compounds to generate new substances [21,22].

R22 and superheated water vapor are mixed in an adiabatic tubular reactor for the cracking reaction, and the main chemical reactions that occur are as follows:

Main chemical reaction equations:

The reaction temperature is 750–790 °C.

$$2\mathrm{C}\mathrm{H}\mathrm{C}\mathrm{l}{\mathrm{F}}_2\to {\mathrm{C}\mathrm{F}}_2={\mathrm{C}\mathrm{F}}_2+2\mathrm{H}\mathrm{C}\mathrm{l}$$

(4)

The free radical reaction mechanism is given below:

$$\mathrm{C}\mathrm{H}\mathrm{C}\mathrm{l}{\mathrm{F}}_2\to :{\mathrm{C}\mathrm{F}}_2+\mathrm{H}\mathrm{C}\mathrm{l}$$

(5)

$$2:{\mathrm{C}\mathrm{F}}_2\to {\mathrm{C}\mathrm{F}}_2={\mathrm{C}\mathrm{F}}_2$$

(6)

Side reaction equations:

The entire reaction process begins with the generation of difluorocarbene radicals through the cleavage of R22. This reaction is a chain-initiated process, marking the start of the overall cleavage reaction. As difluorocarbene is highly reactive, the reaction chain continues to propagate.

$${\mathrm{C}}_2{\mathrm{F}}_4+:{\mathrm{C}\mathrm{F}}_2\to {\mathrm{C}}_3{\mathrm{F}}_6$$

(7)

$${\mathrm{C}}_3{\mathrm{F}}_6+:{\mathrm{C}\mathrm{F}}_2\to {\mathrm{C}}_4{\mathrm{F}}_8$$

(8)

$${\mathrm{C}}_3{\mathrm{F}}_6+\mathrm{H}\mathrm{C}\mathrm{l}={\mathrm{C}}_3{\mathrm{F}}_6\mathrm{H}\mathrm{C}\mathrm{l}$$

(9)

$$\mathrm{C}\mathrm{H}\mathrm{C}\mathrm{l}{\mathrm{F}}_2+{\mathrm{H}}_2\mathrm{O}\to \mathrm{C}\mathrm{O}+2\mathrm{H}\mathrm{F}+\mathrm{H}\mathrm{C}\mathrm{l}$$

(10)

Cracking reactions occur at high conversion rates and often lead to the formation of various fluorine-containing high boilers. These high boilers result from cascade reactions. In this reaction system, water vapor serves as a diluent and heat carrier, helping to prevent the formation of high boilers.

This study infers that the thermal cracking reaction of R22 in water vapor occurs through a free radical mechanism involving diflocarbine. This overall reaction mechanism consists of three key stages: chain initiation, chain growth, and chain termination. Given that diflocarbine is highly reactive and the free radical mechanism is complex, this study focuses on the primary products and a series of by-products generated from one reaction, based on the composition of substances detected during actual production. The cracking reaction model includes nine components and five free radical reactions. A kinetic model for the cracking reaction was developed using the reaction kinetic data that has been validated by the Aspen Plus reactor model [23]. The parameters of the constructed reaction kinetic model are detailed in

Table 1. Cracking reaction kinetic model parameters.

Name	Reaction	A	Ea (kJ/mol)
1	CHClF2→CF2 + HCl	4.8045 × 1010 (s−1)	46.43
2	2:CF2→C2F4	3.592424 × 106 (L/mol∙s)	161.99
3	C2F4 +:CF2→CF2CF2CF2	4.43225 × 106 (L/mol∙s)	13.67
4	:CF2 + H2O→CO + HF	3.0148 × 1013 (L/mol∙s)	104.69
5	2C2F4→CF2CF2CF2CF2	3.133126 × 106 (L/mol∙s)	233.13

.

3. Numerical Simulation

3.1. Numerical Modeling

This simulation studies a tubular cracking reactor used to produce TFE through the water vapor cracking of R22. This reactor is designed as a tubular cracking furnace for this process. The model of the cracking reactor was constructed using Space Claim, based on data provided by the company’s production team [24]. The inlet piping of the reactor features a three-way design, with one inlet for the R22 feedstock, one inlet for water vapor, and one outlet for the product. The diameter of the inlet pipe for R22 is 90 cm, while the inlet for water vapor has a diameter of 125 cm. The outlet pipe, which carries the gas mixture generated after cracking, has a diameter of 150 cm. Additionally, the overall thickness of the piping is set at 5 cm. The final geometrical model is illustrated in Figure 1.

In the Ansys Fluent database, the substances involved in the reaction process are first selected. The database is customized to include the physical parameters of any component that is not already listed and then add the materials. Next, a volumetric reaction is setup in the component transfer model. The reaction equation and kinetic parameters are inputted, and the boundary conditions are established. For the inlet, mass-flow-inlet boundary conditions are used. Specifically, R22 and water vapor are transported at flow rates of 1100 kg/h and 1500 kg/h, respectively. The reactor temperature is set to 750 °C, and for the outlet, a pressure-outlet boundary condition is applied [25,26].

3.2. Meshing

Meshing is performed on the model created in the first step. The inlet and outlet surfaces, as well as the internal and external surfaces, are named to identify the walls and piping. A triangular mesh is chosen as the unit of delineation, with a physical preference for computational fluid dynamics (CFD) to capture the curvature by default and a solver preference for Fluent. Finally, a preliminary mesh is generated by clicking on the “Generate” button. In the reactor model, there is a T-shaped structure; thus, the meshes at the inlet of the R22, the inlet of the water vapor, and the outlet of the mixture are refined to improve mesh accuracy [27]. The three surfaces are selected based on the geometry, adjusted accordingly, and regenerated (Figure 2).

3.3. Grid Quality Check and Independence Verification

Mesh quality significantly affects computational accuracy and stability. Key factors that contribute to mesh quality include node distribution, smoothness, and cell shape. The Mesh Metrics option allows users to view information about these metrics to assess the overall mesh quality [28,29]. After generating the mesh, checking its quality is essential. Once meshing is completed in Fluent Meshing, this study employs a rigorous mesh quality assessment system to validate the computational domain systematically. By switching to the solving mode, the geometric properties of the mesh are evaluated across multiple dimensions, examining key metrics such as mesh orthogonality, aspect ratio, and warping. In this analysis, two inlets are set to different speeds. The final calculation results demonstrate the mesh’s convergence. The evaluation of the computed data reveals that the minimum mesh volume after meshing the cracked reactor is a positive value (6.232724 × 10⁻¹²), indicating there are no negative volumes in the mesh. This ensures the stability of numerical calculations. Furthermore, the mesh orthogonality qualities are all greater than 0.15, with the minimum mesh quality being 0.4724078. This indicates that the grids possess good orthogonal properties, which enhances the computational accuracy (

Table 2. Grid quality check results.

Name	Results
Minimum cell quality	0.4724078
Minimum volume (m3)	6.232724 × 10−12
Maximum volume (m3)	1.236794 × 10−07
Total volume (m3)	7.308784 × 10−03
Minimum face area (m2)	1.994277 × 10−10
Maximum face area (m2)	1.236794 × 10−07

).

Before conducting the simulation calculations, it is essential to perform a grid independence verification. This process ensures that the computational grid is appropriately divided so that the grid density does not significantly affect the simulation results [30]. In this study, five sets of structured grids with varying densities (ranging from 70,000 to 400,000 grids) were created. By incrementally refining the grid and monitoring the trends of key parameters, we determined the optimal grid density that strikes a balance between computational accuracy and resource consumption. The thermal cracking reaction of R22 in water vapor is simulated under consistent operating conditions. Grid independence is verified by ensuring that the variation rate of each parameter at the outlet of the reaction tube does not exceed 1%. Numerical calculations are performed using a steady-state solver, employing the Simple algorithm to address the pressure–velocity coupling, and utilizing an enhanced wall function to accurately capture the flow characteristics near the wall surface. From the data presented in

Table 3. Mesh independence validation results.

Name	70,000	160,000	320,000	350,000	400,000
Total temperature	738.461565	740.298447	748.654935	750.399895	750.395779
Turbulence intensity	86.5872697	115.228765	120.840708	121.160956	120.628291
Outlet mass fraction of H2O	0.56327218	0.52470527	0.50495464	0.50625613	0.50700590
Outlet mass fraction of CHClF2	0.43672781	0.47529473	0.49504535	0.49374386	0.49299409

, it is evident that increasing the number of grids to 320,000 results in minimal effects on individual parameters. This indicates that the grid convergence meets the expected criteria, and the variations in physical quantities at critical monitoring points stabilize. After conducting the independence test, we selected 32,000 grid numbers for the subsequent calculations of the cracking reaction, considering the balance between computational load and accuracy. These quality indices exceed the standard values recommended by Ansys Fluent, providing a reliable numerical basis for further flow field analysis and reactor performance evaluation.

3.4. Solution Setup

Simple is selected as the solution method, the residual of energy iteration is set to 1 × 10⁻⁶, and the residual of the remaining iterations is 1 × 10⁻⁴; the inlet boundary conditions are initialized, then the number of iteration steps is set, and finally, the iterative calculations are carried out.

3.5. Cold Simulation

The primary objective of the cold simulation is to analyze and study the flow characteristics and hydrodynamic behavior of the fluid in the reactor without considering heat exchange [31]. In this model, water vapor and air are utilized for the cold-state simulation, where the mixing process occurs within the reactor. The turbulent flow and mixing of the gasses in the reactor significantly influence temperature, pressure, flow rate, density, and the degree of turbulence. Thus, understanding and optimizing the hydrodynamics and transfer characteristics of the reactor can be achieved by adjusting these conditions and observing the changes in the corresponding parameters.

In this study, we simulated the temperature, flow rate, and turbulence model within a reactor model in a cold state and analyzed their respective cloud diagrams. The air inlet temperature was adjusted to vary between 450 °C and 470 °C, while the water vapor inlet temperature ranged from 900 °C to 950 °C to observe changes in the fluid dynamics within the reactor. Additionally, near the actual flow rate conditions, the flow rates of both the R22 and the water vapor inlets were varied between 1000 and 1700 units in order to investigate flow changes in the reactor. For turbulence modeling, we utilized both the k-ω model and the k-ε model, which are commonly used in Ansys Fluent to simulate and predict turbulence characteristics in fluid flow. The k-ε model is commonly used for isotropic turbulence, while the k-ω model is better suited for boundary layer and anisotropic flow cases.

3.6. Thermal Simulation

Due to the complexity of the cracking reaction products and the limitations of the detection means, it is difficult to detect the data in the cracking reactor. In this study, based on the composition of substances detected in actual production, only the main products and a series of by-products generated by a single reaction are considered as relevant reactions, and this cracking reaction model includes nine components and five radical reaction formulae. Additionally, the numerical simulation results of the constructed model are used to study the flow rate, temperature, pressure, and the distribution of the concentration of each component in the thermal cracking reactor of R22 vapor by using the thermal simulation to quantitatively analyze and comprehend the internal flow and heat and mass transfer laws and provide theoretical basis for the operating parameters and structural parameters in the cracking reactor. A theoretical basis for the operating parameters and structural parameters of the cracking reactor is provided.

The water vapor thermal cracking reactor for R22 has two inlets: one for R22 and another for water vapor. The flow rate for the R22 inlet is approximately 1100 kg/h, while the flow rate for the water vapor inlet is around 1500 kg/h. Both raw materials enter the reactor in the gaseous form [32]. As the feedstocks come from different inlets and have varying flow rates upon entering the reactor, as well as the reactions occurring inside it, these factors influence the flow rate within the reactor itself.

3.7. Reactor Optimization

After conducting cold- and thermal-state simulations of the reactor model, we analyzed the distribution depicted in the cloud diagram and considered optimization options for the reactor. This optimization can be categorized into two main aspects. First, we examined the inlet flow rate. By adjusting the parameters of the two inlet flow rates of the reactor in the software, we expanded the range based on the actual flow rate and selected two inlet flow rates within the range of 0.1 kg/s to 0.5 kg/s. This allowed us to analyze the distribution of components at the outlet and identify the optimal flow rate range. Second, we considered modifications to the structure of the two inlets to enhance their uniformity after mixing. For the three-way inlet reactor used in this simulation, the central vortex created in the mixing channel increases the contact area between reactants, leading to improved performance. Conversely, a T-type inlet reactor, functioning in an engulfing flow state, can enhance the interfacial contact area of the reactants through a three-dimensional vortex structure. This significantly improves mixing, reaction efficiency, and results in a higher and more uniformly distributed product concentration. Additionally, the +-type inlet can achieve better mixing and reaction yield in flow states featuring a central vortex, while the Y-type inlet offers adjustments to its angle based on the original design. In conclusion, both the T-type and three-way inlet reactors can provide effective mixing under specific operational conditions, but the best choice depends on the specific application and circumstances. During the simulation, by modeling the reactor with different inlet structures in the software, we focused on the distribution characteristics of key parameters within the reactor, which included the spatial distributions of the fluid density field, the concentration field of the two reaction components, and the temperature field.

To quantitatively assess the mixing effect, point data for each parameter were analyzed with ANOVA using statistical methods, where the variance value (${\sigma }^2$) was calculated using the formula:

$${\sigma }^2={\displaystyle \frac{\sum {\left(x_i-\mu \right)}^2}{N}}$$

(11)

where x_i is the sampling point data, $\mu$ is the mean value, and N is the total number of sampling points. The smaller the variance value, the more uniform the distribution of parameters and the better the mixing effect [33].

4. Results and Discussion

The cloud diagram in Ansys Fluent is a visualization tool used to represent the results of fluid dynamics simulations. It illustrates the distribution of various physical quantities, such as velocity, pressure, and temperature, using different colors. This allows for a clearer and more intuitive understanding of the fluid flow characteristics in space and the variations in these physical quantities.

4.1. Cold Simulation Data

The following is an analysis of the reactor cold simulation data.

From Figure 3, it is evident that the maximum turbulence flow in the reactor gradually increases as the temperatures of the two inlets rise. This increase is primarily due to the impact of temperature on the density and viscosity of the fluid. As temperature increases, the density and viscosity of the gas tend to decrease, which enhances the kinetic energy of the fluid and promotes turbulence development. Furthermore, the rise in temperature alters the fluid density, directly affecting the Reynolds number (Re). An increase in the Reynolds number is typically associated with the onset of turbulence.

From Figure 4, it is clear that the reactor inlet exhibits a higher flow velocity compared to the overall velocity once the fluid enters the reactor. As the flow velocity at the two inlets increases, the maximum turbulence within the reactor also rises. This phenomenon can be attributed to the differences in flow velocities at the various inlets, which create shear at the fluid interfaces. An increase in flow velocity enhances this shear effect, thereby promoting the formation and development of turbulent vortices. Additionally, an elevated flow velocity directly increases the Reynolds number of the fluid. A higher Reynolds number means that the two gas phases are distributed more uniformly and achieve heat transfer equilibrium more easily.

From Figure 5, it is evident that the maximum turbulence flow case for the k-ω model is higher than that of the k-ε model when two different turbulence models are compared. The k-ω model is more effective at predicting the anisotropy of turbulence in the studied reactor, particularly in the near-wall region and areas of separated flow. Additionally, in complex flow cases—such as those involving multiple turbulence sources or changes in flow direction—the k-ω model is better equipped to capture the intricacies of the flow, leading to a higher predicted turbulence intensity.

In summary, factors such as temperature, pressure, flow rate, density, and the degree of turbulence significantly impact the turbulent flow and mixing processes of the gasses within the reactor model.

4.2. Thermal Simulation Data

The following is an analysis of the reactor’s thermal state simulation data.

The cloud diagram shown in Figure 6 illustrates the temperature and turbulence within the reactor. After the two inlet substances enter the reactor and mix to reach the cracking temperature, a cracking reaction occurs. This reaction is endothermic, meaning it absorbs heat, which initially causes a decrease in the overall temperature. However, a side reaction may occur, and since its reaction rate is relatively high, it can generate localized exothermic heat, leading to an increase in temperature in certain areas. Additionally, due to the inhomogeneity of flow within the reactor, localized hot spots may develop where the concentration of reactants is high, further increasing the temperature. As the reaction takes place, the intensity of turbulence within the reactor increases. Once the side reaction or primary reaction concludes, the overall turbulence intensity decreases.

The cloud diagram in Figure 7 illustrates the distribution of components within the reactor. The two inlet substances entering the reactor are R22 and water vapor. Once inside the reactor and at the cracking temperature, a cracking reaction occurs. As the reaction progresses, the concentrations of the two reactants gradually decrease, while TFE, the main product, is continuously produced. This entire reaction occurs rapidly. Additionally, due to the reactivity of diflocarbine, a series of side reactions can lead to the formation of various by-products.

Figure 8 illustrates the distribution of components at the reactor’s final outlet following the cracking reaction. After processing the data from all the grid points on the outlet plane, the concentration of each component is summarized in

Table 4. Reactor outlet component content.

Name	C2F4	C3F6	C4F8	CHClF2
Mass fraction	0.42316867	0.10547342	5.17392 × 10−14	0.143611705
Name	CO	H2O	HCl	HF
Mass fraction	0.032717123	0.166559308	0.095178699	0.033291074

.

The table presents a comparison between the actual product contents from the plant and the simulation results. It shows that the final product contents from the simulation, conducted under the same reaction conditions, fall within the range observed in industrial production. This confirms the accuracy and reliability of the reactor model used for producing TFE through the thermal cracking of R22 in water vapor, as discussed in this paper. Consequently, the numerical simulation results from the constructed model can be utilized to examine various factors such as flow rate, temperature, pressure, and the concentration distribution of each component in the thermal cracking reactor using R22 and water vapor. This allows for the quantitative analysis and understanding of the internal flow, as well as the heat and mass transfer characteristics. Ultimately, it provides a theoretical basis for determining the operating and structural parameters of the cracking reactor.

4.3. Reactor Optimization Data

The first step is to optimize the reactor inlet flow rate, and the results of the simulation data are as follows. From Figure 9, it can be seen that the optimal flow rate of R22 inlet is 0.2–0.3 kg/s, and the optimal flow rate of water vapor inlet is 0.4–0.5 kg/s.

For the original model of the three-way inlet reactor used in this simulation, the central vortex presented in the mixing channel was able to expand the contact area between the reactants. Another common type of T-type inlet reactor can increase the interfacial contact area of the reactants through the three-dimensional vortex structure, which significantly improves the degree of mixing and reaction, with higher concentration and uniform distribution of the products. The ten-type can also achieve better mixing effect and reaction yield under the flow pattern with central vortex, while the Y-type is based on the original model after adjusting the angle. Therefore, it can be assumed that both the T-type reactor and the three-way inlet reactor can provide good mixing under specific operating conditions, but which one is better may depend on the specific application and operating conditions. In some cases, the T-type reactor may perform better in terms of mixing effectiveness due to its structural characteristics. Therefore, in this study, the reactor inlet structure was optimized and the simulation results are presented below. Cold simulations of the constructed reactor with four different inlet configurations were carried out and the internal structure and distribution of each configuration is shown in Figure 10.

The overall internal data of the above four different structures of the reactor are exported and processed, and the final results are compared in Figure 11.

Through comparative analysis, it is evident that changes in the inlet structure significantly impact the flow field distribution and mixing characteristics within the reactor. In particular, the parameter distribution of the original model exhibits pronounced gradient characteristics, leading to poor mixing. Although the +-type inlet model enhances radial mixing, axial mixing remains inadequate. In contrast, the T-type inlet structure shows the smallest variance in density and component concentration, and its temperature mixing performance is comparable to that of the Y-type structure, which also displays minimal variance. Overall, the data clearly demonstrate the superiority of the T-type inlet structure in promoting the mixing of reactants. Therefore, it is advisable to consider modifying the inlet structure to T-type in the actual production process to improve the mixing of incoming components and facilitate more effective reactions.

5. Conclusions

In this study, we first developed physical, reaction, mathematical, and geometrical models of the cracking reactor. Then, we conducted both analytical setup and solution setup on the constructed model to ensure the accuracy and reliability of the simulation. Next, we performed a mesh quality check and irrelevance verification for the model. After completing the irrelevance verification, we selected the most suitable number of meshes and confirmed that the quality of the final mesh was adequate. This ensured that the impact of the constructed model on the simulation results remained within a reasonable range, thus laying a solid foundation for subsequent simulations of both the cold and thermal states. Finally, our focus shifted to simulating and optimizing the cracking reaction, in which we systematically executed two core tasks: cold-state simulation and thermal -state simulation. Through the cold simulation, we analyzed how temperature, pressure, flow rate, density, and turbulence significantly influence the turbulent gas flow and mixing process within the reactor model. The thermal simulation further explored the effects of the actual reaction process on internal temperature, turbulence, component distribution, outlet component content, and other parameters resulting from changes in the reaction process. Based on our findings, we optimized the inlet flow rate and inlet structure for the reactor’s operating conditions. Our findings indicate that the optimal inlet flow rates are 0.2–0.3 kg/s for R22 and 0.4–0.5 kg/s for water vapor. Additionally, we found that modifying the inlet structure to a T shape achieved the best fluid mixing during the actual production process.

We performed the hydrodynamic simulations of a thermal cracking reactor using Ansys Fluent. This included a range of rigorous analytical setups for model construction and optimization that deepened our understanding of the cracking reaction mechanism. The results provide a more efficient and sustainable method for studying the thermal cracking reactor model for R22 and water vapor, which not only carries out a series of rigorous analytical setups and model constructions and optimizations but also deepens our understanding of the cracking reaction mechanism. Thus, this method provides a scientific basis and theoretical support for achieving the efficient and controllable operation of the cracking process and lays a solid foundation for the subsequent industrialization applications.