Multi-objective optimization method for a bundle of detachable multi-stream wound tubes reactor

Abstract

A multi-objective optimization method for a detachable multi-flow wound tube bundle reactor belongs to the field of reactor optimization design. The reactor employs either a sleeve-type front-end tube box – straight-through multi-flow wound tube bundle – sleeve-type rear-end tube box structure or a sleeve-type front-end tube box – U-shaped multi-flow wound tube bundle structure. This method is based on parametric model generation, automatic mesh generation, and computational fluid dynamics simulation of multi-zone, multi-fluid, multi-phase, and multi-component transport and reaction. It uses multiple process parameters, multiple geometric parameters, and controllable lumped parameters of reaction kinetics as design variables, with operational safety and stability indicators as constraints. Optimization is performed on parameters such as shell-side pressure drop, temperature uniformity, and reactor compactness as multiple objectives to obtain the Pareto optimal solution set. Finally, the optimized structure, process parameters, and performance prediction results are confirmed. This method improves the design accuracy of the reactor's process and structure.

Description

Technical Field

[0001] This invention belongs to the field of reactor design and simulation optimization technology, specifically involving an intelligent evaluation and optimization of heat transfer, pressure drop, and reaction performance of a detachable multi-stream wound tube bundle reactor based on parametric modeling, computational fluid dynamics simulation, and multi-objective optimization algorithms. Background Technology

[0002] Detachable multi-stream wound tube reactors combine the high heat transfer efficiency of shell-and-tube heat exchangers with the excellent reaction performance of fixed-bed reactors, and have been widely used in industries such as petrochemicals, energy and power, ammonia synthesis, Fischer-Tropsch synthesis, methanol synthesis, and alkane dehydrogenation. These reactors typically have catalyst particles packed in the shell side, with the reaction medium flowing through the catalyst bed to undergo a chemical reaction. Simultaneously, heat is removed or supplied through the heat exchange medium in the tube side, allowing for temperature control and suppressing hot spots or runaway reactions.

[0003] Existing wound tube bundle reactors mostly adopt welded or integral non-removable structures, with the tube bundles fixedly connected to the shell. This leads to difficulties in catalyst loading and replacement, inconvenient cleaning and maintenance, and high operating and maintenance costs. Furthermore, it is difficult to flexibly adjust the reactor structure and operating conditions when facing different reaction conditions or load changes.

[0004] Furthermore, the overall operational performance of this type of reactor, such as shell-side pressure drop, heat transfer efficiency, reaction conversion rate, and temperature distribution uniformity, is typically influenced by a combination of factors, including the structural parameters of the wound tube bundle (such as the tube bundle arrangement, winding method, tube diameter, and number of layers), operating parameters (such as flow rate, temperature, and pressure), and catalyst properties (when the catalyst is a coating sprayed onto the outer wall of the tube, the catalytic parameters include the type of catalytic coating on the outer wall, the loading of active components, and the coating thickness; when the catalyst is packed with shell-side particles, the catalytic parameters include the particle size, particle shape, and packing porosity of the shell-side particles; there are no catalytic parameters when the reaction does not require a catalyst). These parameters exhibit complex nonlinear coupling relationships, making it difficult to achieve coordinated optimization of multiple performance indicators using empirical methods or single-parameter adjustments, thus increasing the design cycle and experimental costs. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a multi-objective optimization method for a detachable multi-strand wound tube bundle catalytic reactor. This method achieves convenient assembly, disassembly, and maintenance of the reactor through a modular structure, and optimizes the process through fully automated simulation to balance conflicting performance objectives such as pressure drop, heat transfer, and reaction uniformity, thereby comprehensively improving the reactor's engineering applicability and overall performance.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] Firstly, a detachable multi-strand wound tube bundle reactor is provided. Specifically, it includes the following two structures:

[0008] a. Sleeve-type front-end tube box - straight-through multi-strand spiral tube bundle - sleeve-type rear-end tube box structure, or

[0009] b. Sleeve-type front-end tube box - U-shaped multi-strand wound tube bundle structure.

[0010] For structure a:

[0011] The outer diameter of the rear tube sheet is smaller than the inner diameter of the reactor shell and is inserted into the shell along with the wound tube bundle and fixedly connected to the base of the shell wall. The outer diameter of the front tube sheet is larger than the inner diameter of the reactor shell and is fixed to the flange of the reactor shell.

[0012] For structure b:

[0013] The number of tubes in two adjacent layers of winding is the same, and they are connected one-to-one by a U-shaped return tube section at the bottom of the tube bundle. The inlet and outlet ends are fixed on the same tube sheet.

[0014] Both of the above-mentioned tube sheets are provided with annular grooves for separating multi-flow cavities, and sealing gaskets are provided in the annular grooves; corresponding partition baffles are provided in the reactor head, which are inserted into the annular grooves and cooperate with the sealing gaskets to divide the tube sheet space into multiple independent tube-pass fluid channel cavities.

[0015] Secondly, a multi-objective optimization method for the above-mentioned reactor structure is provided.

[0016] This method constructs a fully automated design process from parameter input to solution output, specifically including the following steps:

[0017] S1. Input reactor key performance design constraints. These include the lower limit of conversion, the upper limit of shell-side pressure drop, and temperature safety limits for reaction safety.

[0018] S2. Define the design variable space, including geometric parameters (including straight pipe section length (L), transition section height (S), number of winding layers (I), winding section height (H), winding radius of each layer R(i), starting angle T(i), number of winding turns M(i), number of pipes N(i), and pipe diameter (do, di) and shell inner diameter Dk), process operation parameters (operating pressure, flow rate, inlet temperature, etc.), and catalytic parameters.

[0019] S3. Based on the input design variable combinations, key performance indicators are obtained through parametric modeling and CFD simulation; this includes the following sub-steps:

[0020] S31. Parametric Modeling: Based on the geometric parameters in the design variable combination, a three-dimensional model of the reactor winding tube bundle is automatically generated; a multi-layer spiral winding tube bundle is generated using the centerline scanning method, wherein the centerline is composed of straight tube segments, spiral segments and smooth transition curves between the two, and the winding directions of adjacent layers alternate; at the same time, the outer shell is automatically generated and configured with a multi-flow structure, and the tube side is divided into multiple independent fluid channels by adding annular baffles.

[0021] S32. Automatic CFD Simulation: Based on the generated geometric model, it automatically configures mesh parameters, physical property parameters and boundary conditions, and calls the solver to complete coupled computational fluid dynamics simulation of multi-region, multi-field, multi-phase, and multi-component transport and reaction;

[0022] S33. Automatically extract key performance indicators such as total shell-side pressure drop, heat transfer efficiency, velocity distribution uniformity, reaction rate distribution uniformity, reaction conversion rate, and temperature extreme values ​​from the simulation results.

[0023] S4. A multi-objective genetic algorithm is used for global iterative optimization. Under the constraints of the lower limit of reaction conversion rate, the upper limit of shell-side pressure drop, and the temperature safety limit, the collaborative objective function of multi-objective optimization is to minimize the total shell-side pressure drop, maximize the total heat transfer coefficient, minimize the shell-side velocity distribution deviation coefficient, minimize the reaction rate distribution non-uniformity, and maximize the reaction conversion rate of key components. In each iteration, the design variable combination is regenerated, and the key performance indicators are obtained. The calculation results are fed back to the algorithm for objective optimization, achieving convergence of the multi-objective Pareto optimal solution set.

[0024] S5. Output the Pareto optimal solution set that satisfies the constraints. Each solution corresponds to a set of design variable combinations and their performance prediction results. Based on specific engineering requirements, the solution set with the highest performance priority can be selected as the final design result.

[0025] Specifically, step S31 includes the following steps, where for each layer i, i is 1 to 1.

[0026] For the sleeve-type front tube box - straight-through multi-strand wound tube bundle - sleeve-type rear tube box structure:

[0027] a. Based on the winding radius R(i), winding height H, and number of turns M(i), a spiral is created using a spiral generation algorithm. The central axis of the spiral is the Z-axis, and the spiral is wound in the positive direction of the Z-axis. For odd-numbered layers (i.e., when i is odd), the winding direction is counterclockwise upward, and for even-numbered layers (i.e., when i is even), the winding direction is clockwise upward.

[0028] b. Calculate the coordinates of each endpoint:

[0029] End point P_he of the leading spiral:

[0030]

[0031] Endpoint P_le of the rear spiral:

[0032]

[0033] Front-end straight pipe segment endpoint P_hu1:

[0034]

[0035] Front straight pipe segment endpoint P_hu2:

[0036]

[0037] Endpoint P_lu1 of the rear straight pipe section:

[0038]

[0039] Endpoint P_lu2 of the back-end straight pipe section:

[0040]

[0041] Where (M(i)-int(M(i))) means taking the fractional part of M(i); T(i) is the starting angle.

[0042] c. Create an upper transition spline curve or serpentine curve between the endpoint of the front spiral P_he and the endpoint of the front straight pipe P_hu1; create a lower transition spline curve between the endpoint of the rear spiral P_le and the endpoint of the rear straight pipe P_lu1.

[0043] d. Constrain the transition spline curves at both ends to maintain a tangent relationship with the corresponding spiral line and straight pipe segment, respectively, and combine them to form a complete single pipe central axis;

[0044] e. Using the aforementioned central axis as the path, scan the circular outline with di as the inner diameter and do as the outer diameter to generate a three-dimensional solid model of a single wound tube;

[0045] f. Based on the number of tubes N(i) in the i-th layer, perform a circular array with the Z-axis as the center to generate all the winding tubes in that layer;

[0046] g. Create a cylindrical shell solid model that surrounds all layers of the wound tube bundle based on the inner diameter Dk of the shell; set the shell-side inlet and outlet to radial and the tube-side inlet and outlet to axial; during modeling, divide different layers into multiple streams; add annular baffles at the corresponding interlayer positions of the orifice plate to divide the interior of the head into multiple independent cavities, using a nested structure, with each cavity corresponding to a tube-side inlet / outlet of a stream.

[0047] For the sleeve-type front-end tube box - U-shaped multi-strand wound tube bundle structure:

[0048] h. Determine the connection endpoints between adjacent layers:

[0049] For the i-th and (i+1)-th layers of the spiral pipe, after generating the spiral lines, transition sections, and straight pipe sections for each layer, the coordinates of the endpoints of the straight pipe sections for both layers are obtained:

[0050] Endpoint of layer i:

[0051]

[0052] Endpoint of layer i+1:

[0053]

[0054] i. Calculate the U-shaped connection plane:

[0055] The plane satisfies: contains and It intersects with the reactor's central axis Z-axis.

[0056] This plane can be determined by the following vectors:

[0057] Connection vector:

[0058] Axial vector:

[0059] j. Construct the center path of the U-shaped tube:

[0060] From the above-mentioned plane and Initially, an arc is generated on each side, gradually changing the flow direction from axial to radial. A connecting path is then created between the two bends, which can be a straight line segment or a spline curve.

[0061] k. Apply the following geometric relationships to the generated U-shaped path: the first bend is tangent to the i-th layer of straight pipe segment, the second bend is tangent to the (i+1)-th layer of straight pipe segment, and the intermediate connecting segment maintains tangential continuity with the bend. This results in a smooth and continuous U-shaped center path.

[0062] 1. To avoid spatial intersections between different flow streams or connecting pipes, the algorithm uses an axially staggered layer arrangement. Specifically, in step S4, the objective function for optimization is:

[0063] Total pressure drop in the shell side f1(x) = ΔP(x)

[0064] The heat transfer efficiency is minimized by taking a negative value: f2(x) = -η(x)

[0065] Shell-side velocity standard deviation f3(x) = σ_v(x)

[0066] The standard deviation of the reaction rate is f4(x) = σ_r(x).

[0067] The conversion rate is minimized by taking a negative value: f5(x) = -X(x)

[0068] Constraints:

[0069] Temperature safety constraints: The maximum shell-side temperature T_max(x) must not exceed the process safety upper limit T_max_s, g1(x)=T_max(x)-T_max_s≤0;

[0070] Pressure drop constraint: The total pressure drop ΔP(x) in the shell side shall not exceed the maximum allowable value ΔP_max, g2(x)=ΔP(x)-ΔP_max≤0;

[0071] Conversion rate constraint: The methanol conversion rate X(x) shall not be lower than the lower limit X_min required by the process, and g3(x) = X_min - X(x) ≤ 0;

[0072] Within the feasible region Ω of the decision variables, find x to minimize the objective vector F(x) = [f1(x), f2(x), f3(x), f4(x), f5(x)], and satisfy g1(x) ≤ 0, g2(x) ≤ 0, g3(x) ≤ 0.

[0073] The beneficial effects of this invention are as follows: the annular baffle divides the interior of the end cap into multiple independent cavities, which can realize independent flow channels for multiple streams without the need for complex internal partitions. While maintaining the detachability of the wound tube bundle, it simplifies the geometric construction and mesh generation, and improves the computational efficiency.

[0074] The entire process from modeling and simulation to optimization is scripted and automated, enabling rapid batch simulation and multi-objective optimization, which significantly shortens the design cycle compared to traditional manual iterative design. Attached Figure Description

[0075] Figure 1 This is a schematic diagram of the overall process of the multi-objective optimization method for the detachable multi-strand wound tube bundle reactor of the present invention.

[0076] Figure 2 This is a schematic diagram of the straight-through multi-strand wound tube bundle of the present invention.

[0077] Figure 3 This is a schematic diagram of the structure of the sleeve-type front-end tube box, straight-through multi-strand wound tube bundle, and sleeve-type rear-end tube box of the present invention.

[0078] Figure 4 This is a schematic diagram of the U-shaped multi-strand wound tube bundle structure of the present invention.

[0079] Figure 5 This is a schematic diagram of the sleeve-type front-end tube box-U-shaped multi-strand wound tube bundle structure of the present invention.

[0080] Figure 6 This is a schematic diagram of the geometric parameters of a single wound tube spiral of the present invention.

[0081] Figure 7 This is a schematic diagram of the geometric parameters of the multilayer wound tube bundle of the present invention.

[0082] The components are: 1. Rear end tube sheet, 2. Front end tube sheet, 3. Cylinder wall base, 4. Annular groove, 5. Dividing baffle, 6. Tube pass 1 inlet, 7. Tube pass 2 inlet, 8. Tube pass 1 outlet, 9. Tube pass 2 outlet, 10. Tube pass U inlet, 11. Tube pass U outlet. Detailed Implementation

[0083] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of protection of this invention.

[0084] Example 1: Sleeve-type front-end tube box - straight-through multi-strand spiral tube bundle - sleeve-type rear-end tube box structure:

[0085] like Figure 2 As shown, the outer diameter of the rear tube sheet 1 is smaller than the inner diameter of the reactor cylinder and is inserted into the cylinder along with the wound tube bundle and fixedly connected to the cylinder wall base 3. The outer diameter of the front tube sheet 2 is larger than the inner diameter of the reactor cylinder and is fixed to the reactor cylinder flange.

[0086] like Figure 3 As shown, each tube sheet is provided with annular grooves 4 for separating multi-flow cavities, and a sealing gasket is provided in the annular groove; a partition baffle 5 is correspondingly provided in the reactor head, the partition baffle is inserted into the annular groove and cooperates with the sealing gasket to divide the tube box space into two independent tube-side channel cavities, namely tube side 1 and tube side 2, and each is provided with an independent tube side 1 inlet 6, tube side 1 outlet 8, tube side 2 inlet 7 and tube side 2 outlet 9.

[0087] Example 2: Sleeve-type front-end tube box - U-shaped multi-strand wound tube bundle structure:

[0088] like Figure 4 As shown, the number of tubes in two adjacent layers of winding tubes is the same, and they are connected one-to-one by a U-shaped return tube section at the bottom of the tube bundle. The inlet and outlet ends are fixed on the same tube sheet.

[0089] like Figure 5 As shown, the tube box structure is the same as in Embodiment 1. The tube side is separated by the cooperation of the tube sheet annular groove and the end cap partition plate, and a tube side U inlet 10 and a tube side U outlet 11 are respectively provided.

[0090] Example 3: Multi-objective optimization for a methanol catalytic oxidation reactor:

[0091] S1. Key performance requirements for the reactor: lower limit of methanol conversion rate, upper limit of shell-side pressure drop, and upper limit of reactor temperature.

[0092] S2. Define the design variable space. The design variable space includes the reactor's geometric parameters (straight pipe length L, transition section height S, number of winding layers I, winding section height H, winding radius of the i-th layer R(i), starting angle of the i-th layer T(i), number of winding turns of the i-th layer M(i), number of tubes of the i-th layer N(i), outer diameter of the winding tube do, inner diameter of the winding tube di, inner diameter of the shell Dk), process operating parameters (inlet temperature, mass flow rate, reactant concentration, and operating pressure of the tube and pipe sides), and catalyst particle parameters.

[0093] S3. Based on a predefined design variable space, the system automatically completes the reactor geometric modeling, simulation model configuration, and numerical calculation process by calling a 3D modeling platform and computational fluid dynamics simulation software that support application programming interfaces (APIs) through parametric program scripts. Specifically, this includes the following steps:

[0094] S31. Parametric geometric modeling. Read the geometric parameters from the design variable combinations, and for each layer i (i = 1 to 5), perform the following operations:

[0095] a. Based on the winding radius R(i), winding height H, and number of turns M(i), a spiral is created using a spiral generation algorithm. The central axis of the spiral is the Z-axis, and winding is performed along the positive Z-axis direction. Odd-numbered layers (i being odd) use a counter-clockwise upward winding direction, while even-numbered layers (i being even) use a clockwise upward winding direction. At the end of the spiral segment, leave a 90° forward angle to allow space for the transition section.

[0096] b. Calculate the coordinates of each endpoint:

[0097] End point P_he of the leading spiral:

[0098]

[0099] Endpoint P_le of the rear spiral:

[0100]

[0101] Front-end straight pipe segment endpoint P_hu1:

[0102]

[0103] Front straight pipe segment endpoint P_hu2:

[0104]

[0105] Endpoint P_lu1 of the rear straight pipe section:

[0106]

[0107] Endpoint P_lu2 of the back-end straight pipe section:

[0108]

[0109] Where (M(i)-int(M(i))) means taking the fractional part of M(i); T(i) is the starting angle.

[0110] c. Create an upper transition spline curve (or serpentine curve) between the endpoint of the front helix P_he and the endpoint of the front straight pipe segment P_hu1; create a lower transition spline curve between the endpoint of the rear helix P_le and the endpoint of the rear straight pipe segment P_lu1.

[0111] d. Constrain the transition spline curves at both ends to maintain a tangent relationship with the corresponding spiral line and straight pipe section, and combine them to form a complete single pipe center axis.

[0112] e. Using the aforementioned central axis as the path, scan the circular outline with di as the inner diameter and do as the outer diameter to generate a three-dimensional solid model of a single wound tube.

[0113] f. Based on the number of tubes N(i) in the i-th layer, perform a circular array with the Z-axis as the center to generate all the winding tubes in that layer.

[0114] g. Create a cylindrical solid model of the shell surrounding all layers of the wound tube bundle, based on the shell's inner diameter Dk. Set the shell-side inlet and outlet to radial (i.e., shell-side fluid flows in and out radially from the shell sidewall), and the tube-side inlet and outlet to axial (i.e., tube-side fluid flows in and out axially). During modeling, divide different concentric layers into multiple flow streams. Add annular baffles at the corresponding interlayer positions of the orifice plate to divide the interior of the head into multiple independent cavities using a nested structure, with each cavity corresponding to a tube-side inlet / outlet of a flow stream.

[0115] After generating the geometric model, the output is a file in a CFD solver-compatible format (.stl format), such as... Figure 7 As shown.

[0116] S32. Automatic CFD Simulation Configuration and Calculation. Background Mesh Generation: The background mesh size is selected as twice the thickness of the wound pipe wall ((do - di) / 2), and the local refinement level is set to (1 2). Based on the selected process operation parameters and catalytic parameters in the design variable space, the necessary physical property parameter files for simulation are automatically generated and written, including the density, specific heat capacity, thermal conductivity, and dynamic viscosity of the fluid medium; lumped kinetic parameters of the chemical reaction, etc. After completing the above configuration, the script program automatically calls the solution module of the computational fluid dynamics software to perform the calculation.

[0117] S4. Further, step 4 specifically involves,

[0118] Decision variable vector: x = [x1, x2, ..., x n ], where x includes structural parameters (number of winding layers I, winding radius of each layer R(i), number of turns per layer M(i), number of tubes per layer N(i), outer diameter of winding tube do, inner diameter of winding tube, etc.), process parameters (shell-side inlet flow rate, temperature, component concentration, flow rate of each stream heat exchange medium and inlet temperature, etc.) and catalyst particle parameters (particle diameter dp, porosity ε, etc.).

[0119] Objective function (converted to a minimization form):

[0120] Total pressure drop in the shell side f1(x) = ΔP(x)

[0121] Heat transfer efficiency (minimizing the negative value): f2(x) = -η(x)

[0122] Shell-side velocity standard deviation f3(x) = σ_v(x)

[0123] The standard deviation of the reaction rate is f₄(x) = σ_r(x).

[0124] Conversion rate (minimize the negative value): f5(x) = -X(x)

[0125] Constraints:

[0126] Temperature safety constraint: The highest shell-side temperature T_max(x) obtained from simulation must not exceed the process safety upper limit T_max_s.

[0127] Pressure drop constraint: The total pressure drop ΔP(x) in the shell side must not exceed the maximum allowable value ΔP_max.

[0128] Conversion rate constraint: The methanol conversion rate X(x) must not be lower than the lower limit X_min required by the process.

[0129] The above constraints can be uniformly expressed as:

[0130] g1(x) = T_max(x) - T_max_s ≤ 0

[0131] g2(x) = ΔP(x) - ΔP_max ≤ 0

[0132] g3(x) = X_min - X(x) ≤ 0

[0133] In summary, the mathematical formulation of the optimization problem is: within the feasible region Ω of the decision variables, find x to minimize the objective vector F(x) = [f1(x), f2(x), f3(x), f4(x), f5(x)], and satisfy g1(x) ≤ 0, g2(x) ≤ 0, g3(x) ≤ 0.

[0134] Create an automated execution script to act as a bridge between the optimization algorithm and the computational fluid dynamics simulation software, and execute the following process:

[0135] The system reads the current design variable combination x generated by the optimization algorithm and dynamically generates a CFD simulation input file corresponding to x based on a parameterized template. It automatically calls the CFD solver to complete the multiphysics simulation calculation under this parameter combination on the computing cluster. After simulation convergence, the script automatically parses and calculates the five objective function values ​​f1(x) to f5(x) corresponding to the design, as well as the constraint function values ​​T_max(x), ΔP(x), and X(x), from the result file. Subsequently, these performance data and constraint violation values ​​are fed back to the optimization algorithm.

[0136] The NSGA-II algorithm was set to a population size of 50 and a maximum number of generations of 500. During the optimization process, the script automatically modeled, simulated, and evaluated multiple combinations of design variables. After approximately 500 iterations, the algorithm converged, yielding a set of Pareto optimal solutions. All solutions in this set satisfied the performance constraints and reflected the interrelationships between multiple objectives, including pressure drop, heat transfer efficiency, temperature uniformity, and reaction conversion rate.

[0137] S5. Output the Pareto optimal solution set. This solution set contains multiple non-dominated solutions, each corresponding to a set of design variable combinations and performance parameters that satisfy the constraints. It can be filtered according to specific engineering requirements. For example, when improving reaction conversion rate is the primary objective, the solution with the higher conversion rate can be selected as the final optimization result, provided other constraints are satisfied. Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-objective optimization method for a detachable multi-strand wound tube bundle reactor, characterized in that, Includes the following steps: S1. Input reactor key performance design constraints, including lower limit of conversion rate, upper limit of shell-side pressure drop, and temperature safety limit; S2. Define the design variable space, including geometric parameters, process operating parameters, and catalytic parameters; S3. Based on the input design variable combinations, key performance indicators are obtained through parametric modeling and CFD simulation; this includes the following sub-steps: S31. Parametric Modeling: Based on the geometric parameters in the design variable combination, a three-dimensional model of the reactor winding tube bundle is automatically generated; a multi-layer spiral winding tube bundle is generated using the centerline scanning method, wherein the centerline is composed of straight tube segments, spiral segments and smooth transition curves between the two, and the winding directions of adjacent layers alternate; at the same time, the outer shell is automatically generated and configured with a multi-flow structure, and the tube side is divided into multiple independent fluid channels by adding annular baffles. S32. Automatic CFD Simulation: Based on the generated geometric model, it automatically configures mesh parameters, physical property parameters and boundary conditions, and calls the solver to complete coupled computational fluid dynamics simulation of multi-region, multi-field, multi-phase, and multi-component transport and reaction; S33. Automatically extract key performance indicators such as total shell-side pressure drop, heat transfer efficiency, velocity distribution uniformity, reaction rate distribution uniformity, reaction conversion rate, and temperature extreme values ​​from the simulation results. S4. A multi-objective genetic algorithm is used for global iterative optimization. Under the premise of satisfying the lower limit of reaction conversion rate, the upper limit of shell-side pressure drop and the temperature safety limit, the cooperative objective function of multi-objective optimization is to minimize the total shell-side pressure drop, maximize the total heat transfer coefficient, minimize the shell-side velocity distribution deviation coefficient, minimize the reaction rate distribution non-uniformity and maximize the reaction conversion rate of key components. In each iteration, the design variable combination is regenerated to obtain the key performance indicators. The calculation results are fed back to the algorithm for target optimization, so as to achieve convergence of the multi-objective Pareto optimal solution set. S5. Output the Pareto optimal solution set that satisfies the constraints. Each solution corresponds to a set of design variable combinations and their performance prediction results. Based on specific engineering requirements, select the solution set with priority to target performance as the final design result.

2. The multi-objective optimization method for the detachable multi-strand wound tube bundle reactor according to claim 1, characterized in that: The geometric parameters are: straight pipe section length, transition section height, number of winding layers, winding section height, winding radius and starting angle of each layer, number of winding turns, number of pipes, inner / outer diameter of the winding pipe, and inner diameter of the outer shell; The process operating parameters are operating pressure, flow rate, and inlet temperature; When the catalyst is a coating sprayed onto the outer wall of the tube, the catalytic parameters include the type of catalytic coating on the outer wall of the tube, the loading of the active component, and the coating thickness. When the catalyst is packed with shell-side particles, the catalytic parameters include the shell-side particle size, particle shape, and packing porosity; when no catalyst is added, there are no catalytic parameters.

3. The multi-objective optimization method for the detachable multi-strand wound tube bundle reactor according to claim 2, characterized in that: The reactor's spiral tube bundle structure is either a sleeve-type front tube box - a straight-through multi-strand spiral tube bundle - a sleeve-type rear tube box structure or a sleeve-type front tube box - a U-shaped multi-strand spiral tube bundle structure.

4. The multi-objective optimization method for the detachable multi-strand wound tube bundle reactor according to claim 3, characterized in that: In the sleeve-type front tube box - straight-through multi-strand wound tube bundle - sleeve-type rear tube box structure, the outer diameter of the rear tube sheet is smaller than the inner diameter of the reactor shell and is inserted into the base inside the shell and fixedly connected with the wound tube bundle as a whole. The outer diameter of the front tube sheet is larger than the inner diameter of the reactor shell and is fixed with the flange of the reactor shell.

5. The multi-objective optimization method for the detachable multi-strand wound tube bundle reactor according to claim 3, characterized in that: In the sleeve-type front-end tube box-U-shaped multi-strand wound tube bundle structure, the number of tubes in adjacent two layers of wound tubes is the same, and they are connected one-to-one by a U-shaped return tube section at the bottom of the tube bundle. The inlet and outlet ends are fixed on the same tube sheet.

6. The multi-objective optimization method for a detachable multi-strand wound tube bundle reactor according to claim 4 or 5, characterized in that: The tube sheet is provided with annular grooves for separating multiple flow channels, and sealing gaskets are installed in the annular grooves. A corresponding partition baffle is installed in the reactor head. The partition baffle is inserted into the annular groove and cooperates with the sealing gasket to divide the tube box space into multiple independent tube-pass fluid channel cavities.

7. The multi-objective optimization method for a detachable multi-strand wound tube bundle reactor according to claim 1, characterized in that: In step S31, for each layer i, i is from 1 to I: For the sleeve-type front tube box - straight-through multi-strand wound tube bundle - sleeve-type rear tube box structure: a. Based on the winding radius R(i), winding height H, and number of turns M(i), a spiral is created using a spiral generation algorithm. The central axis of the spiral is the Z-axis, and the spiral is wound in the positive direction of the Z-axis. For odd-numbered layers (i.e., when i is odd), the winding direction is counterclockwise upward, and for even-numbered layers (i.e., when i is even), the winding direction is clockwise upward. b. Calculate the coordinates of each endpoint: End point P_he of the leading spiral: ； Endpoint P_le of the rear spiral: ； Front-end straight pipe segment endpoint P_hu1: ； Front straight pipe segment endpoint P_hu2: ； Endpoint P_lu1 of the rear straight pipe section: ； Endpoint P_lu2 of the back-end straight pipe section: ； Where (M(i)-int(M(i))) represents taking the decimal part of M(i); T(i) is the starting angle; c. Create an upper transition spline curve or serpentine curve between the endpoint of the front spiral P_he and the endpoint of the front straight pipe P_hu1; create a lower transition spline curve between the endpoint of the rear spiral P_le and the endpoint of the rear straight pipe P_lu1. d. The two ends of the constrained rear transition spline curve are respectively tangent to the corresponding helical line and straight pipe section, and combined to form a complete single pipe center axis; e. Using the central axis as the path, scan the circular outline with di as the inner diameter and do as the outer diameter to generate a three-dimensional solid model of a single wound tube; f. Based on the number of tubes N(i) in the i-th layer, perform a circular array with the Z-axis as the center to generate all the winding tubes in that layer; g. Create a cylindrical shell solid model that surrounds all layers of the wound tube bundle based on the inner diameter Dk of the shell; set the shell-side inlet and outlet to radial and the tube-side inlet and outlet to axial; during modeling, divide different layers into multiple streams; add annular baffles at the corresponding interlayer positions of the orifice plate to divide the interior of the head into multiple independent cavities, and use a nested structure so that each cavity corresponds to a tube-side inlet / outlet of a stream.

8. The multi-objective optimization method for a detachable multi-strand wound tube bundle reactor according to claim 1, characterized in that: In step S31, the following applies to the sleeve-type front-end tube box-U-shaped multi-strand wound tube bundle structure: h. Determine the connection endpoints between adjacent layers: For the i-th and (i+1)-th layers of the spiral pipe, after generating the spiral lines, transition sections, and straight pipe sections for each layer, the coordinates of the endpoints of the straight pipe sections for both layers are obtained: Endpoint of layer i: ； Endpoint of layer i+1: ； i. Calculate the U-shaped connection plane: The plane includes endpoints. and And the plane intersects the reactor's central axis Z-axis; This plane is defined by the following vectors: Connection vector: ; Axial vector: ; j. Constructing the center path of the U-shaped pipe From the above-mentioned plane and Initially, an arc is generated for each bend, which gradually changes the flow direction from axial to radial; a connecting path is generated between the two bends, using a straight line segment or a spline curve. k. Apply the following geometric relationships to the generated U-shaped path: the first bend is tangent to the straight pipe segment of the i-th layer, the second bend is tangent to the straight pipe segment of the (i+1)-th layer, and the intermediate connecting segment maintains tangential continuity with the bend; thus obtaining a smooth and continuous U-shaped center path; l. To avoid spatial intersections between different streams or connecting pipes, the algorithm uses an axially staggered layer arrangement.

9. The multi-objective optimization method for a detachable multi-strand wound tube bundle reactor according to claim 1, characterized in that: In step S4, the objective function for optimizing the objective is: Total pressure drop in the shell side f1(x) = ΔP(x); The heat exchange efficiency is minimized by taking a negative value, f2(x) = -η(x); The standard deviation of the shell-side velocity is f3(x) = σ_v(x); The standard deviation of the reaction rate is f4(x) = σ_r(x); The conversion rate is minimized by taking a negative value, f5(x) = -X(x); Constraints: Temperature safety constraints: The maximum shell-side temperature T_max(x) must not exceed the process safety upper limit T_max_s, g1(x)=T_max(x)-T_max_s≤0; Pressure drop constraint: The total pressure drop ΔP(x) in the shell side shall not exceed the maximum allowable value ΔP_max, g2(x)=ΔP(x)-ΔP_max≤0; Conversion rate constraint: The methanol conversion rate X(x) shall not be lower than the lower limit X_min required by the process, and g3(x) = X_min - X(x) ≤ 0; Within the feasible region Ω of the decision variables, find x to minimize the objective vector F(x) = [f1(x), f2(x), f3(x), f4(x), f5(x)], satisfying g1(x) ≤ 0, g2(x) ≤ 0, g3(x) ≤ 0.

Images