Response surface optimization method for air inlet duct of pendulum mill

Abstract

The invention discloses a response surface optimization method for an air inlet duct of a pendulum mill, and belongs to the technical field of design and optimization of pendulum mills. The problem that component abrasion and collection efficiency cannot be well considered due to structural parameters of an existing pendulum mill is solved. According to the method, modeling is carried out on a main machine of the pendulum mill, a flow field model in the main machine is extracted, and after grid division is carried out, the flow field model is imported into Fluent software to be solved; aiming at the pressure loss of the air duct outlet, respectively fixing an influence factor and taking any other two influence factors as variables to obtain a response curved surface for the influence of the pressure loss so as to determine the pressure loss under the influence of each variable influence factor and a remarkable influence factor; all the influence factors are processed according to the same air duct outlet pressure loss mode, so that candidate values of the influence factors are determined; and according to the candidate values of the influence factors, wind field vortex conditions before and after optimization and abrasion to the blades are analyzed, and finally the structure optimization result of the pendulum mill air inlet duct is determined.

Description

Technical Field

[0001] The invention belongs to the technical field of design and optimization of a swing mill and relates to an optimization method for an air inlet duct of a swing mill. Background Art

[0002] During the operation of the swing mill, the movement trajectory of the material particles will be affected by the force from the flow field. The movement trajectory of the material particles in the flow field can be divided into two stages: the first stage is that the material particles start from a static state and are gradually accelerated by the force of the flow field; the second stage is that the material particles move at a uniform speed along the direction of the flow field after reaching a stable speed. In the first stage, the forces on the material particles mainly include gravity and drag. Gravity accelerates the material particles in a vertical downward direction, while the drag force hinders the falling of the material particles. As the speed of the material particles gradually increases, the drag force they are subjected to gradually decreases until it reaches a balance with gravity, at which time the material particles begin to enter the second stage.

[0003] In the second stage, the forces on the material particles in the flow field mainly include virtual mass force, pressure gradient force, Saffman lift, Basset force and Magnus lift. Virtual mass force is the inertial force on the material particles due to the movement of the fluid, which accelerates the material particles along the direction of the flow field. Pressure gradient force is due to the pressure change generated by the fluid during the movement, which causes the material particles to be subjected to a force along the direction of the pressure gradient. Saffman lift is the lift generated by the friction between the material particles and the fluid, which causes the material particles to deviate from the direction of the flow field. Basset force is the vortex shedding caused by the movement of the material particles in the flow field, which causes the material particles to be subjected to a force deviating from the direction of the flow field. Magnus lift is the lift generated by the rotation of the material particles, which causes the material particles to generate a torque deviating from the direction of the flow field in the flow field. In actual production, in order to improve the efficiency and stability of material particle transportation, it is necessary to optimize the effects of these forces. On the one hand, the parameters of the flow field, such as flow velocity, flow direction and fluid properties, can be adjusted to change the forces acting on the material particles. On the other hand, the properties of the material particles, such as shape, size and density, can be improved to reduce the influence of certain forces acting on the material particles.

[0004] Under the combined action of these forces, the movement trajectory of material particles in the flow field becomes complex and diverse. In order to better understand the movement laws of material particles, researchers have conducted in-depth research on the action mechanisms of these forces through experiments and numerical simulations. These research results have been widely used in engineering practice, such as in powder transportation, particle separation, and fluidized bed reactors. After the material particles leave the scraper, they are affected by gravity. Drag Virtual mass force Pressure gradient force Saffman Lift Basset Force Magnus Lift As shown in the following formula (1):

[0005]

[0006] In the formula, m p is the mass of the particle, kg; is the particle acceleration vector, m / s 2 .

[0007] The flow performance of solid particles in the flow field can be reflected by the Stocks number, that is, the Stocks number = inertia force / resistance. The larger the value, the greater the inertia force on the particles or the smaller the resistance, and the better the flow performance of the material particles. The specific form is as follows

[0008]

[0009] In the formula, ρ p is the particle density of the material, kg / m 3 ;d p is the particle diameter of the material, m; u p is the velocity of material particles in the flow field, m / s; u f is the flow field velocity, m / s; D is the diameter of the flow channel, m.

[0010] From the above formula, we can know that the smaller the particle size, the more easily it is affected by the flow field. Combined with the wear of the air duct components of the swing mill, when the wind field is uniform, the higher the wind speed, the fewer particles pass through the guide vanes. However, too high a wind speed may lead to increased energy consumption or coarseness of the material collected by the mill. The current structural parameters of the swing mill cannot take into account the wear of components and collection efficiency. Summary of the invention

[0011] The invention aims to solve the problem that the structural parameters of the current pendulum mill cannot well take into account the wear of components and the collection efficiency.

[0012] A response surface optimization method for an air inlet duct of a swing mill comprises the following steps:

[0013] First, determine the simulation method and process: for the swing mill host, use 3D modeling software to model, extract the flow field model in the host, and use Fluent meshing to mesh the extracted flow field model; import the obtained mesh file into Fluent software, and solve it in Fluent software. During the solution process, simulate the particles with different particle size distributions in the host, and calculate through Fluent-DEM coupling simulation;

[0014] Based on the simulation method and simulation process, the influencing factors to be optimized are determined, including blade angle, number of blades, and volute line type. The influencing factors of the air duct are used as optimization parameters, and the pressure loss at the air duct outlet, the uneven coefficient, the main engine energy consumption, and the wear of the air duct blades are used as targets for simulation calculation: the experiment is designed using experimental design software and the test results are statistically processed. For the pressure loss at the air duct outlet, one influencing factor is fixed and the other two influencing factors are used as variables to obtain a response surface for the pressure loss, so as to determine the pressure loss and significant influencing factors under the influence of each variable influencing factor; the uneven coefficient, the main engine energy consumption, and the wear of the air duct blades are processed in the same way as the air duct outlet pressure loss; based on the air duct outlet pressure loss, uneven coefficient, main engine energy consumption, and air duct blade wear corresponding to different influencing factors, the candidate values ​​of the influencing factors are determined;

[0015] Based on the simulation method and simulation process, according to the candidate values ​​of the influencing factors, the wind field vortex conditions before and after optimization and the wear of the blades are analyzed, and finally the structural optimization result of the swing mill inlet duct is determined.

[0016] Furthermore, in the process of meshing the extracted flow field model using Fluent meshing, a partitioned meshing method is used to dynamically partition the interior of the swing mill host, distinguish the rotating motion area from the static flow field area, and mesh the flow field inside the swing mill host.

[0017] Furthermore, the flow field in the main unit of the swing mill is meshed, with the global maximum size set to 30 mm, and the minimum mesh size in the rotating area refined to 1 mm.

[0018] Furthermore, in the process of solving the problem in Fluent software, the k-ε solution model was selected and solved using a coupled algorithm.

[0019] Furthermore, in the process of statistically processing the test results using experimental design software, the experimental design software used was Design-Expert version 13.

[0020] Furthermore, the number of blades in the structure of the air inlet duct of the swing mill was finally determined to be 18.

[0021] Furthermore, the blade angle in the structure of the air inlet duct of the swing mill was finally determined to be 55°.

[0022] Furthermore, the final determined volute line shape in the structure of the swing mill air inlet duct is as follows:

[0023] The volute spiral line is drawn using the four-point unequal-distance square method; the volute is a four-segment arc structure, and the radii of the four arcs of the volute are 1870mm, 1740mm, 1600mm, and 1450mm respectively. The arc radius decreases progressively at intervals of 130mm-140mm-150mm.

[0024] Beneficial effects:

[0025] The present invention determines the optimal structural parameters through response surface experimental design. The optimized wind field will help adjust the particle trajectory, reduce component wear and improve collection efficiency, and achieve a good balance between component wear and collection efficiency, while also reducing energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 Schematic diagram of different volute designs.

[0027] Figure 2 Schematic diagram of the mesh division of the flow field inside the main unit of the swing mill.

[0028] Figure 3 This is the response surface diagram of pressure loss.

[0029] Figure 4 It is the velocity cloud diagram of the middle section of the wind duct.

[0030] Figure 5 This is the velocity cloud diagram of the main engine wind field trajectory.

[0031] Figure 6 It is the response surface diagram of the uneven coefficient R.

[0032] Figure 7 This is the response surface diagram of energy consumption.

[0033] Figure 8 This is the response surface diagram of wear amount.

[0034] Fig. 9 Particle distribution diagram at different blade angles.

[0035] Fig.10 It is the wear cloud map of the guide vane and the corresponding enterprise wear-resistant patch map.

[0036] Fig.11 It is the velocity cloud diagram of wind field trajectory.

[0037] Fig.12 This is a cloud diagram of turbulent kinetic energy of wind field traces.

[0038] Fig.13 It is the particle trajectory velocity cloud diagram.

[0039] Fig.14 It is the particle size content curve before and after optimization. DETAILED DESCRIPTION Specific implementation method one:

[0041] This embodiment is a response surface optimization method for an air inlet duct of a swing mill, which determines the optimal structural parameters through response surface experimental design.

[0042] Response Surface Methodology (RSM) is a multivariate data analysis technique that is widely used in engineering, manufacturing, bioscience, finance and other fields. Through this method, the relationship between multiple input variables and output variables can be studied to find the optimal conditions, improve product performance, reduce costs, shorten the R&D cycle, etc. The core idea of ​​the response surface optimization method is to study the relationship between input variables and output variables by establishing a mathematical model between them. Specifically, through experiments or simulations, the data of input variables and output variables are obtained, and then these data are used to fit a mathematical model that can describe the relationship between input variables and output variables.

[0043] Before going into details, let's first explain one of the factors to be optimized, the volute:

[0044] The volute of volute 1 is a four-section type, designed using the traditional structural square method. During the design, according to the size of the entire machine, a square with a side length of 110mm is constructed with the origin as the center. The four vertices of the square are the centers of the four arcs of the volute. The radius of the arc is reduced by 110mm in equal intervals. The radii of the four sections are 1880mm, 1770mm, 1660mm, and 1550mm respectively. Finally, it is manufactured into a volute through welding. This design method is simple to manufacture and convenient for subsequent processing. However, problems such as powder accumulation blockage at the air duct inlet and volute wear occurred during on-site application, indicating that the wind energy at the volute tail is insufficient. Based on experience, the mill manufacturer uses a one-knife cut method at the volute tail to improve the shrinkage of the volute tail, that is, volute 1 (the company's existing volute structure), such as Figure 1 As shown in (a) in .

[0045] Volute 2: The error between the spiral line drawn by the square method and the logarithmic spiral line is too large. In order to improve the performance of the centrifugal fan, based on the CFturbo software design method, the four-point unequal-distance square method is tried to draw the spiral line of the volute. The radius of the four arcs of volute 1 was adjusted. After many attempts and simulations, the radii of the four arcs were 1870mm, 1740mm, 1600mm, and 1450mm respectively. The arc radius was progressively reduced with a spacing of 130mm-140mm-150mm, and volute 2 was obtained. Figure 1 As shown in (b), the advantage of simple manufacturing process is maintained while improving the shrinkage of the volute tail.

[0046] Volute 3: CFturbo software is a structural design software specially used for pumps and fans. In the software, users can choose different design modules according to their needs to design various types of pumps and fans. For example, the software provides a variety of impeller design tools. Users can choose the appropriate impeller structure for design according to the purpose and performance requirements of pumps and fans. In addition, the software also provides powerful drawing functions. Users can directly draw the structural diagram of pumps and fans in the software, and modify and optimize them to make the design more intuitive and efficient. Once again, CFturbo software is highly flexible. It supports the input and output of multiple data formats, and can be seamlessly connected with other related software, making it convenient for users to exchange data and collaborate on design. At the same time, the software also provides a wealth of parameter settings and adjustment functions. Users can adjust the various performance indicators of pumps and fans according to actual needs to meet different design requirements. In fact, it is possible to design and select optimization for volute line types of any other design in the future. This implementation method is based on Figure 1 The volute line shape shown in (c) is described as the volute 3.

[0047] The main steps of the response surface optimization method of the present invention include:

[0048] (1) Experimental design: According to the principles of experimental design, a set of input variables and their value ranges are selected, and experiments or numerical simulations are carried out.

[0049] (2) Fitting response surface model: Based on experimental or numerical simulation data, a mathematical model is established to describe the relationship between input variables and output variables. Commonly used response surface models include polynomial regression models, neural network models, etc.

[0050] (3) Model verification: The model’s predictive ability is verified to examine its validity and reliability.

[0051] (4) Optimization: By analyzing the response surface model and optimizing the algorithm, the optimal solution is found to achieve the goal of optimizing the output variables.

[0052] In order to further explore the distribution law of the main engine wind field, the blade angle, number of blades, and volute line type of the duct are used as optimization parameters, and the duct outlet pressure loss, uneven coefficient, main engine energy consumption, and duct blade wear are used as objective functions. Referring to the reference parameter range given by the enterprise, the Box-Behnken response surface optimization method is used for experimental design. The experimental factor level table is shown in Table 1.

[0053] Table 1 Experimental factor level table

[0054]

[0055] Set Fluent-DEM simulation parameters:

[0056] For the swing mill host, the model was built in the 3D modeling software SpaceClaim, and the internal structure of the host was finely simplified by using its powerful 3D modeling function. During the simplification process, special attention should be paid to the details of the flow field, such as the rotating scraper and blades, so as to reflect the real flow field inside the host to the greatest extent.

[0057] To ensure the accuracy of the internal flow field model, it is necessary to extract the flow field model inside the main machine. Fluent meshing is used to mesh the extracted flow field model. To ensure the quality of meshing, partition meshing is used to dynamically partition the interior of the swing mill main machine, distinguish the rotating motion area from the static flow field area, and mesh the flow field inside the swing mill main machine. Figure 2 As shown in the figure. The purple area at the bottom represents a flow field area where the blade rotates, while the rest of the mesh represents the static flow field area. In order to ensure that the internal flow field environment can be accurately simulated in the simulation, the mesh of the blade area is encrypted. The structure of the blade area is relatively complex and there are many small characteristic structures. Therefore, the global maximum size is set to 30mm in the mesh division, and the mesh size of the rotating area is encrypted to 1mm at the minimum.

[0058] Import the grid file and set the boundary conditions: Import the grid file obtained in the previous step into the Fluent software. In the Fluent software, set the performance parameters of the host under normal operation provided by the enterprise as shown in Table 2. On this basis, select a suitable k-ε solution model and use the coupled algorithm to solve it.

[0059] Table 2 Main engine flow field simulation parameters

[0060]

[0061] In the air flow field inside the main engine, the interaction between gaseous and solid particles forms a gas-solid two-phase flow coupling phenomenon. In order to simulate the wear of the air duct blades, reasonable boundary conditions are first set in the DEM discrete element software. Set the blade speed to 82rev / min; material density to 2700kg / m3; material feed rate to 15t / h; set material parameters, etc.

[0062] The simulation was performed with particles of different particle size distributions in the main engine, as shown in Table 3. Through the Fluent-DEM coupled simulation calculation, the detailed information of the flow field inside the main engine and the mutual influence between the phases can be obtained.

[0063] Table 3 Particle size distribution

[0064]

[0065] Response surface analysis: After simulation tests, the simulation results of each group are sorted out and shown in Table 4. Among them, reducing pressure loss is an important evaluation indicator in mill production. The ventilation pressure loss of the mill and the pressure loss of each key part reflect the size of the system air volume and the wind speed inside the mill. Under the condition of unchanged structure, large pressure loss means large wind speed and large air volume; small pressure loss corresponds to small wind speed and small air volume. Keeping the pressure loss stable reflects the stability of wind speed and air volume, thereby ensuring the stability of the material layer. The unevenness coefficient of the blade duct outlet reflects the uniformity of the air intake duct introducing the main engine and changing the wind speed. It is also an important measure of the stability of the finished product collection wind field inside the main engine. Many swing mills often have serious component wear problems during use due to design or operation reasons. How to reduce wear and increase the service life of equipment is an issue that pendulum mill operators and managers need to pay attention to.

[0066] Table 4 Experimental design and statistical results

[0067]

[0068] The test results were statistically processed using the experimental design (DOE) software Design-Expert version 13. In addition, to verify the validity of the data, a quadratic polynomial regression model was used for corresponding statistical verification. The fitting regression equations for the duct outlet pressure loss P and uneven coefficient R, main engine energy consumption M, and blade wear D were obtained as follows:

[0069] P=403.91+26.09A+14.36B-101.62C+18.62AB-0.8875AC-1.26BC+20.06A 2 +4.37B 2 +162.51C 2

[0070] R=0.3040-0.0663A-0.0238B-0.08C-0.0050AB+0.0175AC+0.0125BC-0.052A 2 -0.0070B 2 -0.0045C 2

[0071] M=0.8540+0.0163A+0.010B-0.0763C+0.015AB-0.0025AC+0.00BC-0.0733A 2 -0.0808B 2 +0.0318C 2

[0072] D=0.0001+0.0001A+0.00B+4.30E-06C+3.150E-06AB+5.625E-06AC+5.250E-07BC+0.00A 2 -2.795E-06B 2 +0.00C 2

[0073] Among them, A, B, and C represent the blade angle, the number of blades, and the volute linear factors, respectively.

[0074] Table 5 Variance analysis table

[0075]

[0076] ※Indicates that the factors have a significant impact on the experimental index (Pr<0.05)

[0077] The variance analysis of pressure loss and non-uniformity coefficient is shown in Table 5, where the lack of fit term is used to indicate the degree of fit between the model and the test. The lack of fit value is greater than 0.05, indicating that there is no lack of fit factor in the model and the Pr value of the model is less than 0.05, indicating that the experimental design results are highly reliable and the model can fit the actual value well. As shown in Table 5, the factors affecting the size of pressure loss are in the order of volute line type > blade angle > number of blades from high to low. The blade angle and volute line type have a significant effect on the size of pressure loss (P < 0.05). Increasing the blade angle and improving the volute line type are both beneficial to improving the uniformity of the internal airflow of the swing mill. The factors affecting the non-uniformity coefficient of the air duct outlet are in the order of volute line type > blade angle > number of blades from high to low, all of which are significant influencing factors (P < 0.05). Among them, changing the blade angle, the number of blades, and the volute line type are all beneficial to improving the uniformity of the internal airflow of the swing mill.

[0078] The response surface of each factor to pressure loss is shown in Figure 3 , Figure 3(a) is the AB factor response surface, (b) is the BC factor response surface, and (c) is the AC factor response surface. When the volute line shape remains unchanged, the pressure loss at the duct outlet continues to increase with the increase of the blade angle and the number of blades. When the blade angle is 55° and the number of blades is 20, the pressure loss reaches a maximum value of 498.86Pa. When the number of blades remains unchanged, the pressure loss at the duct outlet continues to increase with the increase of the blade angle and the poor volute line shape; when the blade angle is 55° and the volute line shape is 1, the pressure loss reaches a maximum value of 701.94Pa. When the blade angle remains unchanged, the pressure loss at the duct outlet continues to increase with the increase of the number of blades and the poor volute line shape. When the number of blades is 20 and the volute line shape is 1, the pressure loss reaches a maximum value of 689.7Pa. The influence of the volute line shape on the pressure loss is the most significant. With the improvement of the volute line shape, the pressure loss is significantly reduced. Properly increasing the blade angle and improving the volute line shape are beneficial to improving the uniformity of the air flow inside the swing type machine and reducing the pressure loss. At the same time, too many blades will lead to an increase in pressure loss.

[0079] Figure 4 The velocity cloud diagram of the middle section of the duct, where (a) is the original model and (b) is the optimized model diagram. According to the statistical results of the response surface experimental design, the duct outlet pressure loss of the combination of blade angle 55°, blade number 18, and volute 1 is the largest, which is 701.94Pa. The velocity cloud diagram of the middle section of the duct of this combination is shown in Figure 4 As shown in (a). Due to the unreasonable volute line shape, the velocity distribution in the volute air duct is uneven. The velocity at the front end of the volute near the air inlet is large and concentrated. At the same time, since the 55° blade angle greatly increases the wind speed compared to the 50° blade angle, there is a positive correlation between the wind speed and the pressure loss, which leads to a large duct pressure loss. The duct outlet pressure loss of the combination of a blade angle of 50°, 18 blades, and volute 2 is the smallest, which is 380.94Pa. After the volute line shape of this combination is optimized, the velocity distribution in the volute air duct is uniform, which can effectively reduce the pressure loss. At the same time, due to the unreasonable blade angle of 50°, a large area of ​​vortex exists between the blades, as shown in the figure. Figure 4 As shown in (b), it is verified that the volute line shape is a significant factor affecting the pressure loss.

[0080] Figure 5 The velocity cloud diagram of the main engine wind field trajectory, where (a) is the original model and (b) is the optimized model diagram. The velocity cloud diagram of the main engine wind field trajectory with a blade angle of 55°, 18 blades and 1 volute is shown in the figure. Figure 5As shown in (a). The wind entering from the inlet enters the main engine through the guide of the volute and the blades. Driven by the blade, the wind shows a spiral upward trend inside the hood. It is intuitively found that the local wind speed in the volute of this group is too high, which increases the pressure loss, and the wind fails to enter the main engine evenly, effectively and timely. In contrast, the main engine with a blade angle of 50°, 18 blades and a volute 2 combination has a more uniform wind field trace distribution. Figure 5 In (b), the wind entering from the inlet can be evenly delivered to the main engine by the optimized volute, which can improve the utilization rate of air volume and thus improve the grinding efficiency. Secondly, the optimized volute reduces wind resistance and reduces energy consumption.

[0081] The response surface of each factor to the non-uniformity coefficient of the air duct outlet is as follows: Figure 6 As shown, (a) is the AB factor response surface, (b) is the AC factor response surface, and (c) is the BC factor response surface. When the volute line shape remains unchanged, as the blade angle and the number of blades continue to decrease, the unevenness coefficient continues to increase. When the blade angle is 46° and the number of blades is 16, the unevenness coefficient reaches the maximum value of 0.33. When the number of blades remains unchanged, as the blade angle continues to decrease and the volute line shape is poor, the unevenness coefficient continues to increase. When the blade angle is 45° and the volute line shape is 1, the unevenness coefficient reaches the maximum value of 0.42. When the blade angle remains unchanged, as the number of blades continues to decrease and the volute line shape is poor, the unevenness coefficient continues to increase. When the number of blades is 16 and the volute line shape is 1, the unevenness coefficient reaches the maximum value of 0.41. The volute line shape has the most significant effect on the unevenness coefficient R of the air duct outlet, followed by the blade angle and the number of blades. Appropriately adjusting the blade angle, increasing the number of blades and selecting a suitable volute line shape can reduce the unevenness coefficient R of the air duct outlet, thereby improving the uniformity of the air flow inside the swing mill.

[0082] Table 6 Variance analysis table

[0083]

[0084] ※Indicates that the factors have a significant impact on the experimental index (Pr<0.05)

[0085] The variance analysis of host energy consumption and blade wear is shown in Table 6. It can be seen from the table that the factors affecting the unit energy consumption of the host are in the order of volute line type > blade angle > number of blades from high to low. The blade angle and volute line type have a significant effect on the energy consumption (Pr < 0.05). Reducing the blade angle and improving the volute line type are both beneficial to reducing the internal energy consumption of the swing mill. The factors affecting blade wear are in the order of blade angle > number of blades > volute line type from high to low. The blade angle and number of blades have a significant effect on wear (Pr < 0.05). Among them, reducing the blade angle and the number of blades is beneficial to reducing the amount of blade wear.

[0086] The response surface of each factor to the host energy consumption is as follows: Figure 7 As shown, (a) is the AB factor response surface, (b) is the AC factor response surface, and (c) is the BC factor response surface. When the volute line shape remains unchanged, with the continuous increase of the blade angle and the number of blades, the unit energy consumption of the main engine shows a trend of first increasing and then decreasing. When the blade angle is 51° and the number of blades is 18, the unit energy consumption of the main engine reaches the maximum value of 0.87 kJ / kg. When the number of blades remains unchanged, with the continuous increase of the blade angle and the optimization of the volute line shape, the unit energy consumption of the main engine shows a downward trend. When the blade angle is 51° and the volute line shape is 1, the unit energy consumption reaches the maximum value of 0.94 kJ / kg. When the blade angle remains unchanged, with the continuous increase of the number of blades and the optimization of the volute line shape, the unit energy consumption of the main engine shows a downward trend. When the number of blades is 18 and the volute line shape is 1, the unit energy consumption reaches the maximum value of 0.93 kJ / kg. It can be seen from Table 6 that the volute line shape has the most significant effect on the energy consumption of the main engine. It can be seen from the figure that with the improvement of the volute line shape, the energy consumption of the main engine is significantly reduced. Properly increasing the blade angle and improving the volute line shape are beneficial to improving the uniformity of air flow inside the swing type machine and reducing energy consumption.

[0087] The response surface of each factor to blade wear is as follows: Figure 8 As shown, (a) is the AB factor response surface, (b) is the AC factor response surface, and (c) is the BC factor response surface. When the volute line shape remains unchanged, the blade wear increases with the increase of the blade angle and the number of blades. When the blade angle is 55° and the number of blades is 20, the wear reaches the maximum value of 1.84E-04m. When the number of blades remains unchanged, the blade wear increases with the increase of the blade angle and the optimization of the volute line shape. When the blade angle is 55° and the volute line shape is 3, the wear reaches the maximum value of 2.06E-04m. When the blade angle remains unchanged, the blade wear increases with the increase of the number of blades and the optimization of the volute line shape. When the number of blades is 20 and the volute line shape is 3, the wear reaches the maximum value of 1.13E-04m. The blade angle has the most significant effect on blade wear. As can be seen from the figure, as the blade angle decreases, the blade wear decreases significantly. Appropriately reducing the blade angle and the number of blades is conducive to reducing the wear of the blades and increasing the service life of the components.

[0088] In a swing mill, the guide vanes introduce the wind field into the main machine, guide the movement of materials, and improve the crushing and collection efficiency. However, if the angle of the guide vanes is not designed properly, it will cause eddy currents in the wind field, many particles between the blades, and severe wear. When the angle of the guide vanes is set to 50°, Fig. 9As shown in (a), the blade angle design is unreasonable, there are more vortices between the blades, the uniformity is poor, the material keeps rotating between the blades, the friction and collision between the particles and the blades increase, resulting in severe wear. Fig. 9 As shown in (b), after the flow field is optimized, the vortex between the blades is reduced, the flow field is smoother, and the particles between the blades are reduced, which reduces wear. This is of great significance for improving the operating stability of the equipment and extending the service life of the blades. At the same time, due to the reduction in wear, the maintenance cost of the equipment is also reduced accordingly, thereby reducing the overall operating cost of the enterprise.

[0089] In order to reduce the wear of the guide vanes, the first thing that comes to mind is to adjust the angle of the blades. Through research, it is found that adjusting the angle of the guide vanes to 55° can reduce the wear between the blades to a certain extent. It should also be noted that the wear problem of the position where the inner side of the 55° guide vane directly contacts the particles still exists. After many tests and studies, the above adjustment cannot solve the wear problem well, and it will also affect other factors. Therefore, wear-resistant patches are used to solve the wear problem. Fig.10 For the wear cloud map of the guide vanes and the corresponding wear-resistant patches, the company decided to install wear-resistant patches on the inner side of the guide vanes. Fig.10 As shown in the real picture on the right side of the middle, the wear-resistant patch is a composite material with high wear resistance. Installing it on the guide vane can effectively improve the wear resistance of the blade and extend its service life. After the implementation of this method, significant results have been achieved. It not only solves the problem of severe wear on the inner side of the guide vane, but also reduces the company's maintenance costs and improves production efficiency.

[0090] Through response surface analysis, multi-objective optimization based on response surface is carried out under the premise of minimizing pressure loss and non-uniformity coefficient, and the optimal volute and blade structure combination data are obtained. The optimal solution is volute 2, 18 blades, and a blade angle of 55°. The following is a comparative study of the company's existing mill solution (volute 1, blade angle 50°, 18 blades) and the optimized solution. The whole machine is simulated by combining the DPM multiphase flow analysis technology in Fluent, and the collection efficiency of 200 mesh (75μm) finished particles is studied. Among them, the collection efficiency of the pendulum mill has an important impact on its competitiveness in the field of ultrafine powder production. The level of this indicator directly reflects the competitive strength of the pendulum mill in the industry. The collection efficiency M can be calculated according to formula (3):

[0091]

[0092] Where N e is the mass flow rate of finished particles collected at the outlet, kg / s; N iis the total fine particle mass flow rate at the inlet, kg / s. Specifically, the collection efficiency is the ratio of the mass of the finished particles collected from the crushed fine particles by the mill to the mass of the total fine particles.

[0093] Fig.11 The velocity cloud diagram of the wind field of the whole machine, where (a) is the original model and (b) is the optimized model diagram. The velocity cloud diagram of the wind field of the original model shows a relatively messy distribution, the wind field flow is not stable enough, and there are large-scale eddies, which increases energy consumption, such as Fig.11 (a), and after optimization, that is, Fig.11 (b) The wind field trace cloud diagram becomes more orderly and stable. The wind flows inward along the air duct under the guidance of the blades. The rotation of the grinding roller blade will form a vortex in the middle area of ​​the main engine. The wind field in the main air duct rotates upward along the inner wall of the hood barrel. The uniform flow field significantly improves the vortex phenomenon and reduces energy consumption.

[0094] Fig.12 The turbulent kinetic energy cloud diagram of the wind field trace of the whole machine, where (a) is the original model and (b) is the optimized model diagram. In the original model, the local turbulent kinetic energy of the blade area and the classifier classification area is large, and there are obvious vortex areas and uneven distribution, which significantly reduces the airflow speed. Fig.12 (a). The model after structural optimization, such as Fig.12 (b), the turbulent kinetic energy between the blades is small, and the vortex area has been significantly improved. The turbulent concentration area in the classification area of ​​the classifier is reduced, and the wind field is evenly distributed, thereby reducing the vortex. The stable wind field is conducive to the transportation of particles, ultimately achieving the goal of reducing energy consumption and improving collection efficiency.

[0095] Fig.13 The particle trajectory cloud diagram shows the trajectory and speed of particles in the wind field. Fig.13 In (a), the uneven wind field causes some particles to have a relatively concentrated and irregular trajectory when approaching the classifier. This unstable motion trajectory reduces the collection efficiency and makes it impossible for the particles to be effectively captured and transported. Optimized model Fig.13 In (b), the wind field is significantly improved and more evenly distributed. This uniform wind field makes the movement of particles orderly and stable, which is conducive to the particles reaching the classifier position accurately.

[0096] Further data sorting of the simulation results showed that the pressure loss of the original model was 2808.53Pa, the outlet wind speed was 33.7m / s, and the pressure loss of the optimized model was 2836.34Pa, the outlet wind speed was 32.93m / s. According to the pressure loss ΔP (Pa) formula: ΔP = K*ρ*V 2 / 2; where K is the resistance coefficient; ρ is the air density, kg / m3; V is the wind speed, m / s. There is a positive correlation between wind speed and pressure loss. The optimized model duct blade is 55°, which increases the wind speed entering the main engine, speeds up the air flow in the main engine, and increases the resistance of the blade area, causing greater pressure loss. It is known that the height of the whole machine is 4.9m. According to formula 2-4, the unit mass energy consumption of the original model and the optimized solution is 1.98kJ / kg and 2.03kJ / kg respectively. Due to the change in the blade angle, the two solutions have caused a change in the inlet duct resistance coefficient. Therefore, the energy consumption of the optimized solution is slightly improved. According to formula 2-5, the power consumption of the whole machine part of the original model and the optimized solution is 37.6kW·h and 38.55kW·h respectively. Further calculation of the collection efficiency shows that the collection efficiency of the finished product (particles within 200 mesh) of the original model is 84%, while the collection efficiency of the finished product of the optimized model is increased to 95.3%, which is 13.4% higher than the original model. That is, under the same level of equipment energy consumption, the equipment's production capacity is greatly improved and the optimization effect is obvious.

[0097] The optimized mill was tested for production, and limestone was selected as the raw material. The BT-9300T laser particle size analyzer was used to compare the particle size of the products. The particle size content curve is shown in Fig.14 The cumulative collection efficiency is shown in Table 7. Compared with the collection efficiency of 200-mesh products, the original design of the company is 89.36%, which is highly consistent with the simulation analysis. The optimized version of the mill is 95.30%, so the data is highly reliable. The optimization scheme can obtain smaller and finer powder size, and the product quality is more stable.

[0098] Table 7 Cumulative collection efficiency

[0099]

[0100] This paper adopts Fluent simulation technology and response surface optimization technology to conduct response surface experimental analysis on the optimization design of the air inlet duct of the swing mill, and studies the pressure loss at the duct outlet, the uneven coefficient of the duct blade outlet, energy consumption, blade wear and finished product collection efficiency. The blade wear process is studied in detail through the coupling of Fluent and DEM, and the characteristics and mechanism of blade wear under different working conditions are discussed, which provides a theoretical basis for optimizing blade design and improving its service life. The following conclusions are drawn:

[0101] (1) The Box-Behnken response surface design method was used to study the influence of blade angle, blade number, and volute line shape on duct outlet pressure loss and duct unevenness coefficient, main engine energy consumption, and blade wear. The response surface analysis results show that for duct outlet pressure loss, blade angle and volute line shape are both significant factors, and the blade number factor is not significant; for the duct outlet unevenness coefficient, the three factors are all significant factors, and the volute line shape is the most significant; for main engine energy consumption, blade angle and volute line shape are both significant factors, and the volute line shape is the most significant, and the blade number factor is not significant; for blade wear, blade angle and blade number are both significant factors, and the volute line shape is not significant; appropriately adjusting the blade angle and optimizing the volute structure can effectively improve the ventilation performance of the mill and improve production efficiency. Considering the manufacturing cost, the optimal solution is 18 blades, a blade angle of 55°, and a volute of 2.

[0102] (2) Verify the effectiveness of the optimized design. By comparing the original model with the optimized solution, the finished product collection efficiency of the original model is 84%, while the finished product collection efficiency of the optimized model is increased to 95.3%, an increase of 13.4% over the original model. The pressure loss of the original model is 2808.53Pa, and the pressure loss of the optimized model is 2836.34Pa, which is slightly higher than the pressure loss of the original model, but it is negligible compared with the improvement of the finished product collection efficiency. This result shows that the optimization design and simulation analysis can realize the wind field control and verification simulation of the mill, and at the same time proves the potential of the research method in improving the production efficiency of the swing mill, providing a feasible solution for equipment optimization in the field of industrial milling.

[0103] The present invention may also have many other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art may make various corresponding changes and modifications based on the present invention, but these corresponding changes and modifications should all fall within the scope of protection of the claims attached to the present invention.

Claims

1. A response surface optimization method for an air inlet duct of a swing mill, characterized in that: The following steps are involved: First, determine the simulation method and process: for the swing mill host, use 3D modeling software to model, extract the flow field model in the host, and use Fluent meshing to mesh the extracted flow field model; import the obtained mesh file into Fluent software, and solve it in Fluent software. During the solution process, simulate the particles with different particle size distributions in the host, and calculate through Fluent-DEM coupling simulation; Based on the simulation method and simulation process, the influencing factors to be optimized are determined, including blade angle, number of blades, and volute line type. The influencing factors of the air duct are used as optimization parameters, and the pressure loss at the air duct outlet, the uneven coefficient, the main engine energy consumption, and the wear of the air duct blades are used as targets for simulation calculation: the experiment is designed using experimental design software and the test results are statistically processed. For the pressure loss at the air duct outlet, one influencing factor is fixed and the other two influencing factors are used as variables to obtain a response surface for the influence of the pressure loss, so as to determine the pressure loss under the influence of each variable influencing factor and the significant influencing factors; the uneven coefficient, the main engine energy consumption, and the wear of the air duct blades are processed in the same way as the pressure loss at the air duct outlet; Determine candidate values ​​of influencing factors based on the duct outlet pressure loss, unevenness coefficient, main engine energy consumption, and duct blade wear corresponding to different influencing factors; Based on the simulation method and simulation process, according to the candidate values ​​of the influencing factors, the wind field vortex conditions before and after optimization and the wear of the blades are analyzed, and finally the structural optimization result of the swing mill inlet duct is determined.

2. The response surface optimization method for the air inlet duct of a swing mill according to claim 1, characterized in that: In the process of meshing the extracted flow field model using Fluent meshing, the partitioned meshing method is used to dynamically partition the interior of the swing mill host, distinguish the rotating motion area from the static flow field area, and mesh the flow field inside the swing mill host.

3. The response surface optimization method for the air inlet duct of a swing mill according to claim 2, characterized in that: The flow field in the main unit of the swing mill is meshed, with the global maximum size set to 30 mm, and the minimum mesh size in the rotating area refined to 1 mm.

4. The response surface optimization method for the air inlet duct of a pendulum mill according to claim 1 is characterized in that: In the process of solving the problem in Fluent software, the k-ε solution model is selected and the coupled algorithm is used for solving the problem.

5. The response surface optimization method for the air inlet duct of a swing mill according to claim 1, characterized in that: In the process of statistical processing of the test results using experimental design software, the experimental design software used was Design-Expertversion 13.

6. A response surface optimization method for an air inlet duct of a swing mill according to any one of claims 1 to 5, characterized in that: The number of blades in the final structure of the air inlet duct of the swing mill is 18.

7. The response surface optimization method for the air inlet duct of a pendulum mill according to claim 6, characterized in that: The blade angle in the final structure of the air inlet duct of the swing mill is 55°.

8. The response surface optimization method for the air inlet duct of a pendulum mill according to claim 7, characterized in that: The final determined volute line shape in the structure of the swing mill air inlet duct is as follows: The volute spiral line is drawn using the four-point unequal-distance square method; the volute is a four-segment arc structure, and the radii of the four arcs of the volute are 1870mm, 1740mm, 1600mm, and 1450mm respectively. The arc radius decreases progressively at intervals of 130mm-140mm-150mm.

Images