Large tube plate quenching simulation heat transfer coefficient calibration method based on particle swarm optimization

Abstract

A large tube plate quenching simulation heat exchange coefficient calibration method based on a particle swarm algorithm comprises the steps that tube plate quenching data are collected, a particle swarm algorithm model is built, a tube plate quenching finite element simulation model is built, quenching heat exchange coefficients are updated through the particle swarm algorithm, and the heat exchange coefficient of a forge piece and air and the heat exchange coefficient of the forge piece and a medium are obtained. A large tube plate quenching simulation heat transfer coefficient is calibrated by using a particle swarm algorithm, based on a finite element simulation technology and on the basis of the deviation between a time-temperature data simulation value and a production value in the quenching process of each part of the tube plate, so that the simulation prediction precision is improved, and the tube plate quenching numerical simulation can effectively guide process optimization and actual production.

Description

Technical Field

[0001] This invention relates to the field of tube sheet quenching technology, and in particular to a method for calibrating the heat transfer coefficient in large tube sheet quenching simulation based on particle swarm optimization algorithm. Background Technology

[0002] The tube sheet of the hydrotreating reactor, as a core forging involved in the conversion of heavy oil, plays a crucial role in supporting the catalyst bed, buffering thermal stress, providing high-pressure sealing, and resisting sulfur and hydrogen. Therefore, the steel used for the tube sheet needs high resistance to high-temperature tempering brittleness, hydrogen corrosion, and weldability. Heat treatment is a critical process to ensure the final product performance meets standards, with quenching playing a vital role in guaranteeing heat treatment performance. However, due to the large size of the tube sheet forgings and the complex heat exchange with the environment, temperature control during quenching is extremely difficult. Controlling the temperature field is a crucial means to ensure the transformation of the internal microstructure of the forging.

[0003] With the development and application of finite element method (FEM) and computer technology, numerical simulation is increasingly widely used in the formulation of heat treatment process parameters. By simulating the heat treatment of tube sheets using FEM, the microstructure, properties, and temperature field changes during quenching can be obtained. However, the consistency between current numerical simulation results and actual production still has many problems, providing only limited reference for process design. The most significant factor is the large error between the heat transfer coefficient of the material and the environment and the actual situation. Most simulations use a fixed constant heat transfer coefficient, while in reality, it is a fluctuating value that changes with the temperature of the material and the medium. Therefore, it is necessary to optimize the calibration technology for the heat transfer coefficient in large tube sheet quenching simulations to improve the accuracy of tube sheet heat treatment simulations. Summary of the Invention

[0004] To address the issue of low accuracy in predicting loads during large tube sheet quenching simulations, this invention provides a method for calibrating the heat transfer coefficient in large tube sheet quenching simulations based on the particle swarm optimization algorithm.

[0005] The technical solution adopted by this invention to solve the above-mentioned technical problems is: a method for calibrating the heat transfer coefficient of large tube sheet quenching simulation based on particle swarm optimization algorithm, comprising the following steps: Step 1: Collect tube sheet quenching data, including water / oil immersion temperature, duration, quenching medium, and temperature-time change data during the quenching process of the tube sheet forging. Then, perform correlation processing on the data, and use the water / oil immersion temperature, duration, and quenching medium of the same forging quenching process as input for the tube sheet forging quenching simulation, and use the time-temperature change during the quenching process as the benchmark for judging the deviation of the simulation results. Step 2: Build the particle swarm optimization (PSO) algorithm model, which includes: S1. Initialize the particle swarm Particle count: Sets the population size N; Position and velocity: each particle i is randomly initialized; Position x i Randomly distributed within the search space; speed v i The initial velocity is set to 0 or a small random value; Individual optimal (p) best The initial optimal position of each particle is its initial position; Global optimal (g) best ): Select the position of the particle with the best fitness from the population; S2, Particle Swarm Iterative Update Through multiple iterations, the velocity and position of each particle are updated; Speed ​​updated to: ; In the above formula, w is the inertia weight, used to balance the global and local search and control the inheritance of particle velocity; c1 and c2 are acceleration constants, used to adjust the weight of individual experience and social experience; r1 and r2 are random numbers in [0,1]. Location updated to: ; When updating the position, check if the particle's position exceeds the search boundary; if it does, correct it. S3. Terminate the iteration after obtaining the optimal solution through updating. Individual optimality: If the current fitness is better than p best Then update ; Global optimality: If the fitness of a certain particle is better than g best Then update g best ; When the maximum number of iterations or g is reached best Stop iterating when the change is less than the threshold. Step 3: Establish a finite element simulation model for tube sheet quenching. First, a model of the tube sheet forging with uniform geometric dimensions after rough machining is established and imported into the finite element simulation software for mesh generation. Then, the tube sheet material and CCT curve are defined, and the forging exit temperature t1, the time from exiting the furnace to entering the quenching medium hoisting time s1, the quenching time s2, and the quenching medium type are set. Step 4: Update the quenching heat transfer coefficient using the particle swarm optimization algorithm. Initialize the heat transfer coefficient w1 between the forging and air, and the heat transfer coefficient w2 between the forging and the medium. Run the finite element simulation software to obtain the time-temperature change data during the quenching process, calculate the fitness, and adjust the particle trajectory based on the individual historical best position and the collective historical best position of the particles through the particle swarm algorithm model built in step two. Iteratively update the heat transfer coefficient, where the heat transfer coefficient w1 corresponds to the simulation of the forging hoisting stage, and the heat transfer coefficient w2 corresponds to the simulation of the forging immersion in water / oil quenching stage. When the fitness reaches the threshold or the maximum number of iterations is reached, the iteration stops, and the heat transfer coefficient w1 between the forging and air and the heat transfer coefficient w2 between the forging and the medium are output.

[0006] Preferably, in step one, tube sheet quenching data is collected via OPC, Modbus protocol, or ACCESS database.

[0007] Preferably, in S1, the population size N is set to 20-50 particles.

[0008] Preferably, in S2, c1=c2≈2.

[0009] According to the above technical solution, the beneficial effects of the present invention are: This invention uses particle swarm optimization and finite element simulation technology to calibrate the heat transfer coefficient of large tube sheet quenching simulation based on the deviation between the simulated and actual values ​​of time-temperature data during the quenching process of various parts of the tube sheet. This improves the accuracy of simulation prediction and ensures that the numerical simulation of tube sheet quenching can effectively guide process optimization and actual production. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0011] A method for calibrating the heat transfer coefficient in large-scale tube sheet quenching simulation based on particle swarm optimization algorithm, the process of which is as follows: Figure 1 As shown, it includes the following steps: Step 1: Collect tube sheet quenching data via OPC, Modbus protocol, and ACCESS database. This includes the water / oil immersion temperature, duration, quenching medium, and temperature-time variation data during the quenching process of the tube sheet forging. Then, perform correlation processing on the data. Use the water / oil immersion temperature, duration, and quenching medium of the same forging quenching process as input for the tube sheet forging quenching simulation, and use the time-temperature variation during the quenching process as the benchmark for evaluating the deviation of the simulation results.

[0012] Step 2: Build the particle swarm optimization (PSO) algorithm model, which includes: S1. Initialize the particle swarm Number of particles: Set the population size N (20-50 particles); Position and velocity: Each particle i is randomly initialized: Position x i Randomly distributed within the search space; speed v i The initial velocity is set to 0 or a small random value; Individual optimal (p) best The initial optimal position of each particle is its initial position; Global optimal (g) best ): Select the position of the particle with the best fitness from the population.

[0013] S2, Particle Swarm Iterative Update The velocity and position of each particle are updated through multiple iterations.

[0014] Speed ​​updated to: ; In the above formula, w is the inertia weight, used to balance the global and local search and control the inheritance of particle velocity; c1 and c2 are acceleration constants, used to adjust the weight of individual experience and social experience, usually c1=c2≈2; r1 and r2 are random numbers in [0,1].

[0015] Location updated to: ; When updating the position, check if the particle's position exceeds the search boundary; if it does, correct it.

[0016] S3. Terminate the iteration after obtaining the optimal solution through updating. Individual optimality: If the current fitness is better than p best Then update ; Global optimality: If the fitness of a certain particle is better than g best Then update g best ; When the maximum number of iterations or g is reached best The iteration stops when the change is less than the threshold.

[0017] Step 3: Establish a finite element simulation model for tube sheet quenching. First, a model of the tube sheet forging with uniform geometric dimensions after rough machining is established and imported into the finite element simulation software for mesh generation. Then, the tube sheet material and CCT curve are defined, and the forging exit temperature t1, the time s1 for hoisting the forging from the furnace to the quenching medium, the quenching time s2, and the type of quenching medium are set.

[0018] Step 4: Update the quenching heat transfer coefficient using the particle swarm optimization algorithm. Initialize the heat transfer coefficient w1 between the forging and air, and the heat transfer coefficient w2 between the forging and the medium. Run the finite element simulation software to obtain the time-temperature change data during the quenching process, calculate the fitness, and adjust the particle trajectory based on the individual historical best position and the collective historical best position of the particles through the particle swarm algorithm model built in step two, and iteratively update the heat transfer coefficient.

[0019] The heat transfer coefficient w1 corresponds to the simulation of the forging hoisting stage, and the heat transfer coefficient w2 corresponds to the simulation of the forging immersion in water / oil quenching stage. Therefore, the same iteration of the tube sheet quenching finite element simulation model can output two heat transfer coefficients simultaneously.

[0020] When the fitness reaches the threshold or the maximum number of iterations is reached, the iteration stops, and the heat transfer coefficient w1 between the forging and air and the heat transfer coefficient w2 between the forging and the medium are output.

[0021] Example 1: Data collection and processing of tube sheet production data.

[0022] Thermocouples and other equipment were installed around the circumference and on the upper and lower surfaces of the tube sheet during the heat treatment process to record production data. Production data such as quenching temperature (850~870℃), hoisting time (153~187s), and quenching medium (oil) of 14 hydrogenated φ7500×300mm tube sheets were collected using OPC, Modbus protocol, and ACCESS database. The material was 16Mn, and its CCT curve was measured by a thermal expansion meter. The above data were used as input for the tube sheet quenching simulation, and the time-temperature data during the quenching process was used as the benchmark for evaluating the load deviation of the simulation output.

[0023] II. Calibration of tube sheet heat transfer coefficient based on particle swarm optimization algorithm model.

[0024] 1. Initialize the particle swarm Particle (heat transfer coefficient w1 between forging and air, heat transfer coefficient w2 between forging and medium) quantity: Set particle population size to 35.

[0025] Position and velocity are randomly initialized for each particle (w1, w2)i.

[0026] Position x i : Random distribution.

[0027] speed v i The initial velocity is set to 0.

[0028] Individual optimal (p) best ): Set the initial optimal position of each particle as its initial position.

[0029] Global optimal (g) best ): Selecting the position of the particle with the best fitness from the population.

[0030] 2. Design of simulation environment for tube sheet quenching.

[0031] Select the hydrogenation tube sheet with the most complete production data as the object, and establish its rough-machined geometric model: φ7500×300mm. Import it into the finite element simulation software with a mesh size of 15. Define the tube sheet material as 16Mn, input its CCT curve, and set the forging exit temperature to 860℃, the time from exiting the furnace to entering the quenching medium to be hoisted to 163s, the quenching time to 1800s, and the quenching medium to be water.

[0032] 3. Iterative updates Run the solver and perform iterative calculations through finite element simulation. For each iteration, update the velocity and position of each particle (w1, w2).

[0033] 4. Update the optimal solution Update individual optimality: if the current fitness is better than p best Then update .

[0034] Update global optimum: If the fitness of a certain particle is better than g best Then update g best .

[0035] Under the condition of calculating the optimal solution, the deviation (fitness) between the simulated temperature field change of the tube sheet quenching and the actual temperature field is calculated. If the fitness does not reach the optimal value, the individual and global optimal solution particles (w1, w2) are updated in the 16Mn tube sheet quenching simulation program to continue the optimization iteration.

[0036] 4. Termination Conditions When the deviation between the simulated and actual time-temperature curves at various points on the tube sheet is less than 5%, the iteration ends, and the heat transfer coefficients w1 between the forging and air and w between the forging and the medium are output in the form of data tables, thus obtaining the curves of w1, w2 and temperature.

[0037] III. Testing and Verification The output heat transfer coefficients w1 and w2 were written into the finite element software to simulate another tube sheet forging with different quenching processes such as furnace exit temperature (870℃) and quenching time (2100s). The deviation of the simulation from the actual production temperature field of each part of the tube sheet was calculated. The experimental results proved that this embodiment can reliably improve the simulation prediction accuracy and ensure that the numerical simulation of tube sheet quenching can effectively guide process optimization and actual production.

Claims

1. A method for calibrating the heat transfer coefficient in large-scale tube sheet quenching simulation based on particle swarm optimization algorithm, characterized in that, Includes the following steps: Step 1: Collect tube sheet quenching data, including water / oil immersion temperature, duration, quenching medium, and temperature-time change data during the quenching process of the tube sheet forging. Then, perform correlation processing on the data, and use the water / oil immersion temperature, duration, and quenching medium of the same forging quenching process as input for the tube sheet forging quenching simulation, and use the time-temperature change during the quenching process as the benchmark for judging the deviation of the simulation results. Step 2: Build the particle swarm optimization (PSO) algorithm model, which includes: S1. Initialize the particle swarm Particle count: Sets the population size N; Position and velocity: Each particle i is randomly initialized; Position x i Randomly distributed within the search space; speed v i The initial velocity is set to 0 or a small random value; Individual optimal (p) best The initial optimal position of each particle is its initial position; Global optimal (g) best ): Select the position of the particle with the best fitness from the population; S2, Particle Swarm Iterative Update Through multiple iterations, the velocity and position of each particle are updated; Speed ​​updated to: ； In the above formula, w is the inertia weight, used to balance the global and local search and control the inheritance of particle velocity; c1 and c2 are acceleration constants, used to adjust the weight of individual experience and social experience; r1 and r2 are random numbers in [0,1]. Location updated to: ; When updating the position, check if the particle's position exceeds the search boundary; if it does, correct it. S3. Terminate the iteration after obtaining the optimal solution through updating. Individual optimality: If the current fitness is better than p best Then update ; Global optimality: If the fitness of a certain particle is better than g best Then update g best ; When the maximum number of iterations or g is reached best Stop iterating when the change is less than the threshold. Step 3: Establish a finite element simulation model for tube sheet quenching. First, a model of the tube sheet forging with uniform geometric dimensions after rough machining is established and imported into the finite element simulation software for mesh generation. Then, the tube sheet material and CCT curve are defined, and the forging exit temperature t1, the time from exiting the furnace to entering the quenching medium hoisting time s1, the quenching time s2, and the quenching medium type are set. Step 4: Update the quenching heat transfer coefficient using the particle swarm optimization algorithm. Initialize the heat transfer coefficient w1 between the forging and air, and the heat transfer coefficient w2 between the forging and the medium. Run the finite element simulation software to obtain the time-temperature change data during the quenching process, calculate the fitness, and adjust the particle trajectory based on the individual historical best position and the collective historical best position of the particles through the particle swarm algorithm model built in step two. Iteratively update the heat transfer coefficient, where the heat transfer coefficient w1 corresponds to the simulation of the forging hoisting stage, and the heat transfer coefficient w2 corresponds to the simulation of the forging immersion in water / oil quenching stage. When the fitness reaches the threshold or the maximum number of iterations is reached, the iteration stops, and the heat transfer coefficient w1 between the forging and air and the heat transfer coefficient w2 between the forging and the medium are output.

2. The method for calibrating the heat transfer coefficient in large tube sheet quenching simulation based on particle swarm optimization as described in claim 1, characterized in that: In step one, tube sheet quenching data is collected via OPC, Modbus protocol, and ACCESS database.

3. The method for calibrating the heat transfer coefficient in large tube sheet quenching simulation based on particle swarm optimization as described in claim 1, characterized in that: In S1, the population size N is set to 20-50 particles.

4. The method for calibrating the heat transfer coefficient in large tube sheet quenching simulation based on particle swarm optimization algorithm as described in claim 1, characterized in that: In S2, c1 = c2 ≈ 2.

Images