Pore ​​modeling method for gas diffusion layer of proton exchange membrane fuel cell

Abstract

The present invention discloses a porosity modeling method for a gas diffusion layer of a proton exchange membrane fuel cell, comprising the following steps: S1. formulating the volume parameters and the number of fiber layers of the gas diffusion layer; S2. selecting the probability distribution type of the target porosity according to the target porosity of each fiber layer, and determining the probability distribution parameters of the probability distribution type; S3. using a script parameterized modeling method, randomly generating the target porosity value of each fiber layer according to the probability distribution characteristics of the target porosity; S4. designing the fiber layer according to the target porosity value, generating fibers according to the fiber model in the fiber layer, and stretching the two ends of the fiber so that the fibers penetrate the fiber layer transversely; S5. performing fiber removal in the fiber layer and calculating the porosity of the fiber layer, and calculating the porosity of the fiber layer by iteration until the target porosity requirement is met; S6. merging the fiber layers after the design of each fiber layer is completed until a gas diffusion layer is formed as a whole. The present invention improves modeling efficiency and operation convenience.

Description

Technical Field

[0001] The invention relates to the technical field of proton exchange membrane fuel cells, and in particular to a modeling method for a fiber pore structure of a gas diffusion layer of a proton exchange membrane fuel cell. Background Art

[0002] The water management problem of proton exchange membrane fuel cells is an important factor that determines the performance of fuel cells. Since the reaction gases in the anode and cathode channels need to pass through the gas diffusion layer to reach the catalyst layer for reaction, the liquid water generated by the cathode catalyst layer after the reaction needs to be discharged to the channel through the gas diffusion layer, otherwise it will block the reaction gas path. Therefore, the drainage performance of the gas diffusion layer has a huge impact on the performance of the fuel cell.

[0003] In order to analyze the drainage performance of the gas diffusion layer, its real structure must be reconstructed first. The material and manufacturing method of the gas diffusion layer make its structure similar to randomly placed cylinders. It is impossible to establish a realistic model manually, and it is also difficult to control its porosity. Therefore, it is necessary to develop a new modeling method to control the porosity of the gas diffusion layer and improve modeling efficiency and ease of operation. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the prior art and propose a script parameterized modeling method with controllable porosity error, which can present the real pore morphology of porous media and improve modeling efficiency and operation convenience.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] A pore modeling method for a proton exchange membrane fuel cell gas diffusion layer comprises the following steps:

[0007] S1. Formulate volume parameters and the number of fiber layers of the gas diffusion layer, wherein the top surface of the fiber layer contacts the bottom surface of the upper fiber layer, and the bottom surface of the fiber layer contacts the top surface of the lower fiber layer;

[0008] S2. Selecting a probability distribution type of a target porosity according to the target porosity of each fiber layer, and determining a probability distribution parameter of the probability distribution type;

[0009] S3. using a script parameterized modeling method to randomly generate a target porosity value of each fiber layer according to the probability distribution characteristics of the target porosity;

[0010] S4. designing the fiber layer according to the target porosity value, generating fibers in the fiber layer according to the fiber model, and stretching both ends of the fibers so that the fibers penetrate the fiber layer transversely;

[0011] S5. performing fiber removal in the fiber layer and calculating the porosity of the fiber layer, and calculating the porosity of the fiber layer by iteration until the target porosity requirement is met;

[0012] S6. After the design of each fiber layer is completed, the fiber layers are combined until they become an integrated gas diffusion layer.

[0013] Furthermore, in step S1, the fiber layer of the gas diffusion layer is in a rectangular parallelepiped shape, and the volume parameters of the fiber layer include the length, width and thickness of the fiber layer.

[0014] Furthermore, in step S2, the probability distribution type of the target porosity includes single-point distribution, uniform distribution, normal distribution or user-defined distribution.

[0015] Furthermore, in step S3, script parametric modeling is based on ANSYS-SpaceClaim scripts and supports Python language modeling by calling the Python standard library random library.

[0016] Furthermore, in step S4, the fiber model adopts a cylindrical model, and the diameter of the bottom surface of the cylinder is smaller than the thickness of each fiber layer.

[0017] Further, the central axis of the cylinder is on the central cross section of the fiber layer, and the fibers generated in the fiber layer are cross-arranged.

[0018] Furthermore, the fibers of the fiber layer are randomly arranged, and two points are randomly generated on the central cross section of the fiber layer, and the two points are used as the centers of two bottom surfaces of the fibers.

[0019] Further, in step S4, both ends of the fiber are stretched outwards to penetrate the fiber layer transversely, and the stretching distance is 1.5 times the maximum dimension of the volume parameter of the fiber layer.

[0020] Furthermore, step S5 specifically includes the following steps:

[0021] S51. Let i be the layer variable, iter be the iteration variable of each fiber layer, iter=0 is the initial state, and the initial porosity of the fiber layer ε0=1;

[0022] S52. For the i-th layer, the target porosity of the fiber layer is ε t [i], one fiber is generated in each iteration, iter=iter+1, the fiber is cut off, and the porosity ε of the fiber layer is updated. iter ;

[0023] S53. Calculation of porosity ε iter Whether the first iteration condition ε is satisfied iter -ε t[i]<η, η is the error limit, if the porosity ε iter If the first iteration condition is not met, return to step S52;

[0024] S54. If the porosity ε iter Satisfy the first iteration condition and calculate the porosity ε iter Whether the second iteration condition ε is met iter -ε t [i]>-η, if the porosity ε iter When the second iteration condition is met, the iteration of the i-th layer is completed, the iteration data of the fiber layer porosity is recorded, and the modeling of the lower fiber layer begins;

[0025] S55. If the porosity ε iter If the second iteration condition is not met, the iter-th iteration is canceled, and the iter-th fiber is generated by custom inserting fibers.

[0026] Further, in step S52, the porosity ε iter The calculation formula is

[0027]

[0028] Where V iter-1 Vf is the pore volume of the fiber layer after the iter-1th iteration, iter is the volume of the fiber removed from the iter-th iteration fiber layer, V iter-1 -Vf iter is the pore volume of the fiber layer after the iterth iteration, and V0 is the initial total volume of the fiber layer.

[0029] Compared with the prior art, the present invention reproduces the random distribution of carbon fibers when manufacturing gas diffusion layers through a random modeling method, solves the problem that manual modeling is difficult to present the true pore morphology of porous media, improves modeling efficiency and operational convenience through script parameterized modeling, and sets the iterative termination condition to regulate the porosity error accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A schematic flow chart of a specific embodiment of a pore modeling method for a gas diffusion layer of a proton exchange membrane fuel cell.

[0031] Figure 2 Schematic diagram of randomly generating two points on the central cross section of the fiber layer.

[0032] Figure 3 A schematic diagram of generating a fiber model with two randomly generated points in a fiber layer as the centers of the circles at the bottom surfaces of both ends of the fiber.

[0033] Figure 4Schematic diagram of the two ends of the stretched fiber in the fiber layer.

[0034] Figure 5 It is a schematic cross-sectional diagram of the central cross section of the fiber layer after the fibers in the fiber layer are cut off.

[0035] Figure 6 Generate a custom fiber area division diagram for the fiber layer.

[0036] Figure 7 Schematic diagram of determining fiber positions for a fiber layer.

[0037] Figure 8 Schematic diagram of the superposition model of fiber layers with target porosity obeying normal distribution.

[0038] Fig. 9 Schematic diagram of the model of the gas diffusion layer that combines various fiber layers.

[0039] Fig.10 Schematic diagram of the actual porosity distribution characteristics of each fiber layer that obeys the normal distribution.

[0040] Fig.11 Schematic diagram of the error distribution characteristics of actual porosity when the target porosity obeys normal distribution.

[0041] Fig.12 Schematic diagram of the model of the gas diffusion layer in which the porosity of each fiber layer is a fixed value of 0.6.

[0042] Fig.13 Schematic diagram of the actual porosity error distribution characteristics when the target porosity is a fixed value. DETAILED DESCRIPTION

[0043] The pore modeling method of the gas diffusion layer of a proton exchange membrane fuel cell of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0044] like Figure 1 As shown, the present invention discloses a random modeling method for the three-dimensional pore structure of a gas diffusion layer of a proton exchange membrane fuel cell, which specifically includes the following steps:

[0045] S1. The volume parameters and number of fiber layers of the gas diffusion layer are determined, the top surface of the fiber layer contacts the bottom surface of the upper fiber layer, and the bottom surface of the fiber layer contacts the top surface of the lower fiber layer.

[0046] The fiber layer of the gas diffusion layer is in the shape of a cuboid, and the volume parameters of the fiber layer include the length, width and thickness of the fiber layer. The length of the fiber layer is Lx, the width of the fiber layer is Ly, the thickness of the fiber layer is Lz, and the total number of fiber layers of the gas diffusion layer is n. The diameter of the carbon fiber is generally 5-20 microns. For the convenience of modeling, it is magnified 1000 times during modeling, that is, millimeters are used as the unit. When further solving after the pre-processing is completed, the size parameters are scaled 1000 times.

[0047] S2. Selecting a probability distribution type of a target porosity according to the target porosity of each fiber layer, and determining a probability distribution parameter of the probability distribution type.

[0048] The probability distribution type can be single point distribution (fixed value), uniform distribution, normal distribution or custom distribution. Single point distribution (fixed value) means that the target porosity of each fiber layer takes the same value c. The density function of single point distribution (fixed value) is Dirac delta function, which is:

[0049]

[0050] The mathematical property of the Dirac delta function is that the probability density of points other than the c value is equal to zero, and the integral over the entire domain is equal to 1. The uniform distribution performs random sampling with equal probability by setting the interval parameter [lb, ub] of the target porosity. The normal distribution performs random sampling by setting the mean μ and standard deviation σ parameters of the target porosity. The custom distribution refers to sampling that conforms to other specific distributions or customizing the target porosity value of each layer according to actual design requirements.

[0051] S3. Using the ANSYS-SpaceClaim script modeling function that supports Python, the target porosity value ε of each layer is randomly generated by calling the Python standard library random library t =[ε1,ε2,...,ε n ] and set a random seed to ensure that each simulation reproduces the same results.

[0052] S4. The fiber layer is designed according to the target porosity value, fibers are generated in the fiber layer according to the fiber model, and both ends of the fibers are stretched so that the fibers penetrate the fiber layer transversely.

[0053] The fiber model adopts a cylindrical model, and the radius of the bottom of the cylinder is set. The bottom diameter of the cylinder is slightly smaller than the thickness of each fiber layer, so that the fiber layer is not cut when the Boolean cutting operation is further performed. The central axis of the cylinder is on the central cross section of the fiber layer, and the fibers generated in the fiber layer are cross-arranged.

[0054] The arrangement of each fiber is determined by two random points, which are randomly generated in the central cross section of the fiber layer, such as Figure 2 As shown, two points A and B are randomly generated on the cross section, and points A and B are used as the centers of the two bottom surfaces of the fiber, as shown in Figure 3 As shown, the connecting line is the central axis, and a fiber can be generated for every two random points. In order to penetrate the fiber layer horizontally, the two ends of the generated cylinder are stretched outwards respectively, and the stretching distance is 1.5 times the maximum size of the fiber layer volume parameter, as shown in Figure 4 After the fiber model is generated, a Boolean cut operation is performed with the fiber layer, and the center cross-sectional view is shown in Figure 5 shown.

[0055] S5. Perform fiber removal in the fiber layer and calculate the porosity of the fiber layer, and iteratively calculate the porosity of the fiber layer until the target porosity requirement is met.

[0056] The design process starts from the first layer and designs one by one. The iterative termination condition is formulated, that is, the error accuracy ε is specified. When the error between the porosity and the target porosity reaches the error accuracy range through iteration, the iteration is stopped and the design of the pore structure of this layer is completed.

[0057] Wherein, step S5 specifically includes the following steps:

[0058] S51. Let i be the layer variable, iter be the iteration variable of each layer of fiber, iter=0 is the initial state, and the initial porosity of the fiber layer ε0=1.

[0059] S52. For the i-th layer, the target porosity of the fiber layer is ε t [i], one fiber is generated in each iteration, iter=iter+1, the fiber is cut off, and the porosity ε of the fiber layer is updated. iter , porosity ε iter The calculation formula is:

[0060]

[0061] Where V iter-1 Vf is the pore volume of the fiber layer after the iter-1th iteration, iter is the volume of the fiber removed from the iter-th iteration fiber layer, V iter-1 -Vf iter is the pore volume of the fiber layer after the iterth iteration, and V0 is the initial total volume of the fiber layer.

[0062] S53. Calculation of porosity ε iter Whether the first iteration condition ε is satisfied iter -ε t [i]<η, where η is the error limit. If the first iteration condition ε is not met iter -ε t[i]<η, return to step S52.

[0063] S54. If the porosity ε iter Satisfy the first iteration condition ε iter -ε t [i]<η, calculate the porosity ε iter Whether the second iteration condition ε is met iter -ε t [i]>-η.

[0064] If the porosity ε iter Satisfy the second iteration condition ε iter -ε t [i]>-η, the iteration of the i-th layer is completed, the porosity iteration data of the fiber layer is recorded, and the modeling of the next layer begins, i=i+1. If i=n+1 is not satisfied, return to step S51; if i=n+1 is satisfied, the modeling of each fiber layer is completed.

[0065] S55. If the porosity ε iter The second iteration condition ε is not satisfied iter -ε t [i]>-η, cancel the iter-th iteration, and generate fibers by custom inserting fibers for the iter-th iteration.

[0066] Define the iterth fiber arrangement direction to be parallel to the diagonal of the central cross section of the fiber layer, such as Figure 6 As shown, there are four distribution forms, upper left D0, lower left D1, lower right D2 and upper right D3. The distribution area is selected by taking the remainder. Let t be the custom fiber insertion order. The remainder is expressed as t%4 in Python programming to ensure that the custom fibers are not concentrated in the same area when inserted continuously.

[0067] The cross section of the fiber layer center is a square, with a side length of L, a fiber model radius of r, and a thickness of 2r to simplify the calculation. In the D0 region, the fiber center axis intersects with the edge, and the distance between the intersection and the lower right corner is h, as shown in Figure 7 As shown, the relationship between the ratio of the total volume of a fiber generated, Pf, and h is:

[0068]

[0069] To ensure that fibers are not generated in other areas,

[0070] In programming, using the formula By properly adjusting h, the position and volume of the last fiber can be controlled. After Boolean operation, the porosity requirements can be met. h can be calculated using the following formula:

[0071]

[0072] In the formula, ε iter-1 Refers to the porosity of the fiber layer after canceling the iter iteration.

[0073] If the calculation result exceeds the upper limit, the upper limit is used as the basis to generate fibers as the iterth iteration, and the process returns to step S53. If the result is lower than the lower limit, the iteration is stopped and the next layer modeling is started, i=i+1. If i=n+1 is not satisfied, the process returns to step S51; if i=n+1 is satisfied, the modeling of each layer is completed.

[0074] S6. After completing the design of all fiber layers, merge the fiber layers through Boolean operation to form an integrated gas diffusion layer.

[0075] Taking the porosity as an example, the mean μ is 0.6, the standard deviation σ is 0.06, the bottom radius of the cylindrical fiber model is 4 mm, the length, width and thickness of each fiber layer are 100 mm, 100 mm and 10 mm respectively, and the error limit is set to ±0.003. The models of each layer are as follows: Figure 8 As shown, after merging Fig. 9 The distribution characteristics of the porosity of each fiber layer are shown in Fig.10 As shown, it obeys the normal distribution, and the error distribution is as follows Fig.11 As shown. Fig.11 It can be seen that the porosity of each fiber layer is within the error range and meets the accuracy requirements.

[0076] The porosity is a single-point distribution (fixed value 0.6) as the target porosity, the bottom radius of the cylindrical fiber model is 4 mm, the length, width and thickness of each fiber layer are 100 mm, 100 mm and 10 mm respectively, and the error limit is set to ±0.003. The modeling results of the method of the present invention are as follows: Fig.12 The distribution of porosity of each fiber layer is shown in Fig.13 As shown. Fig.13 It can be seen that the porosity of each fiber layer is within the error range and meets the accuracy requirements.

[0077] In summary, the present invention reproduces the random distribution characteristics of carbon fiber when manufacturing the gas diffusion layer through a random modeling method, solves the problem that manual modeling is difficult to present the true pore morphology of porous media, improves modeling efficiency and operational convenience through script parameterized modeling, and sets the iterative termination condition to regulate the porosity error accuracy.

[0078] The above description is a detailed description of the preferred feasible embodiments of the present invention, but the embodiments are not intended to limit the scope of the patent application of the present invention. All equivalent changes or modified changes completed under the technical spirit disclosed by the present invention should fall within the patent scope covered by the present invention.

Claims

1. A pore modeling method for a proton exchange membrane fuel cell gas diffusion layer, characterized in that: The following steps are involved: S1. Formulate volume parameters and the number of fiber layers of the gas diffusion layer, wherein the top surface of the fiber layer contacts the bottom surface of the upper fiber layer, and the bottom surface of the fiber layer contacts the top surface of the lower fiber layer; S2. Selecting a probability distribution type of a target porosity according to the target porosity of each fiber layer, and determining a probability distribution parameter of the probability distribution type; S3. using a script parameterized modeling method to randomly generate a target porosity value of each fiber layer according to the probability distribution characteristics of the target porosity; S4. designing the fiber layer according to the target porosity value, generating fibers in the fiber layer according to the fiber model, and stretching both ends of the fibers so that the fibers penetrate the fiber layer transversely; S5. performing fiber removal in the fiber layer and calculating the porosity of the fiber layer, and calculating the porosity of the fiber layer by iteration until the target porosity requirement is met; S6. After the design of each fiber layer is completed, the fiber layers are combined until they become an integrated gas diffusion layer.

2. The pore modeling method of a proton exchange membrane fuel cell gas diffusion layer according to claim 1, characterized in that: In step S1 , the fiber layer of the gas diffusion layer is in a rectangular parallelepiped shape, and the volume parameters of the fiber layer include the length, width and thickness of the fiber layer.

3. The pore modeling method of a proton exchange membrane fuel cell gas diffusion layer according to claim 1, characterized in that: In step S2, the probability distribution type of the target porosity includes single-point distribution, uniform distribution, normal distribution or user-defined distribution.

4. The pore modeling method of a proton exchange membrane fuel cell gas diffusion layer according to claim 1, characterized in that: In step S3, script parametric modeling is based on ANSYS-SpaceClaim scripts and supports Python language modeling by calling the Python standard library random library.

5. The pore modeling method of a proton exchange membrane fuel cell gas diffusion layer according to claim 1, characterized in that: In step S4, the fiber model adopts a cylindrical model, and the diameter of the bottom surface of the cylinder is smaller than the thickness of each fiber layer.

6. The pore modeling method of a proton exchange membrane fuel cell gas diffusion layer according to claim 5, characterized in that: The central axis of the cylinder is on the central cross section of the fiber layer, and the fibers generated in the fiber layer are cross-arranged.

7. The pore modeling method of a proton exchange membrane fuel cell gas diffusion layer according to claim 6, characterized in that: The fibers of the fiber layer are randomly arranged, and two points are randomly generated on the central cross section of the fiber layer, and the two points are used as the centers of the two bottom surfaces of the fibers.

8. The pore modeling method of a proton exchange membrane fuel cell gas diffusion layer according to claim 1, characterized in that: In step S4, both ends of the fiber are stretched outwards to penetrate the fiber layer transversely, and the stretching distance is 1.5 times the maximum dimension of the volume parameter of the fiber layer.

9. The pore modeling method of a proton exchange membrane fuel cell gas diffusion layer according to claim 1, characterized in that: Step S5 specifically includes the following steps: S51. Let i be the layer variable, iter be the iteration variable of each fiber layer, iter=0 is the initial state, and the initial porosity of the fiber layer ε0=1; S52. For the i-th layer, the target porosity of the fiber layer is ε t [i], one fiber is generated in each iteration, iter=iter+1, the fiber is cut off, and the porosity ε of the fiber layer is updated. iter ; S53. Calculation of porosity ε iter Whether the first iteration condition ε is satisfied iter -ε t [i]<η, η is the error limit, if the porosity ε iter If the first iteration condition is not met, return to step S52; S54. If the porosity ε iter Satisfy the first iteration condition and calculate the porosity ε iter Whether the second iteration condition ε is met iter -ε t [i]>-η, if the porosity ε iter When the second iteration condition is met, the iteration of the i-th layer is completed, the iteration data of the fiber layer porosity is recorded, and the modeling of the lower fiber layer begins; S55. If the porosity ε iter If the second iteration condition is not met, the iter-th iteration is canceled, and the iter-th fiber is generated by custom inserting fibers.

10. The pore modeling method of a proton exchange membrane fuel cell gas diffusion layer according to claim 9, characterized in that: In step S52, the porosity ε iter The calculation formula is Where V iter-1 Vf is the pore volume of the fiber layer after the iter-1th iteration, iter is the volume of the fiber removed from the iter-th iteration fiber layer, V iter-1 -Vf iter is the pore volume of the fiber layer after the iterth iteration, and V0 is the initial total volume of the fiber layer.

Images