JP-2026075608-A - A method for predicting the aging and decay of gas diffusion layers based on a combination of offline and online data.

JP2026075608AJP 2026075608 AJP2026075608 AJP 2026075608AJP-2026075608-A

Abstract

[Problem] To provide a method for predicting the aging decay of a gas diffusion layer based on a combination of offline and online methods. [Solution] The gas diffusion layer to be predicted and its initial performance are acquired, offline and online cyclic bench tests for durability aging decay are performed, and the test time is recorded. After completion, the degradation performance of the gas diffusion layer to be predicted is collected, and an offline performance test is performed on each of the samples after the offline aging test and the samples obtained by disassembling the test stack. The average performance of the gas diffusion layer to be predicted at that time is acquired, and the set of aging decay acceleration coefficients from the offline durability test and the online cyclic bench durability test is calculated and acquired, and the average value is calculated by adding them together to obtain the final combination of acceleration coefficients, and then the final acceleration coefficient is obtained by summing the averages. [Selection Diagram] Figure 1

Inventors

焦道寛
▲ハオ▼ 冬
呉志新
▲衛▼ ▲ポン▼男
王睿迪
張妍懿
侯永平
浦及
馬明輝

Assignees

中汽研新能源汽車検験中心（天津）有限公司
中国汽車技術研究中心有限公司

Dates

Publication Date: 20260508
Application Date: 20251016
Priority Date: 20241022

Claims (1)

A method for predicting the aging decay of a gas diffusion layer based on a combination of offline and online methods, To obtain the gas diffusion layer to be predicted and the initial performance of the gas diffusion layer to be predicted, The gas diffusion layer to be predicted is mounted on multiple test stacks of equivalent output and size, Each test stack undergoes a cyclic test of durability aging and decay according to the set driving mode of the actual vehicle, and the cycle test time for each test stack is recorded. After the cycle test is completed, the test stacks after multiple aging and decay periods with different cycle test times are disassembled one by one, the gas diffusion layer to be predicted is extracted, and a performance test is conducted to obtain the first average degradation performance of the gas diffusion layer to be predicted at that time. An offline durability test is performed on the gas diffusion layer to be predicted, the offline durability test time is recorded, and after completion, the second average degradation performance of the gas diffusion layer to be predicted after the adoption of the offline durability test is collected. Based on the second average degradation performance and initial performance of the gas diffusion layer to be predicted, the aging decay width of the gas diffusion layer to be predicted after cycle-mounted durability of each test stack is calculated and obtained. Based on the aging decay width, first average performance of degradation, second average performance of degradation, and initial performance of the gas diffusion layer to be predicted, a set of aging decay acceleration coefficients for offline durability tests and online cycled bench durability tests is calculated and obtained. This includes adding the indices of each term in the set of aging decay acceleration coefficients to their corresponding values, calculating the average value to obtain the final combination of acceleration coefficients, and then performing an average of these additions to obtain the final acceleration coefficients, Obtaining the gas diffusion layer to be predicted and the initial performance of the gas diffusion layer to be predicted is, The measurement indicators for the gas diffusion layer to be predicted include thickness H, tensile strength T, roughness R, contact angle C, planar resistivity E, and air permeability P. If the initial performance of the gas diffusion layer to be predicted is X, X=[H, T, R, C, E, P] (1) The performance of the gas diffusion layer to be predicted after aging and decay includes thickness H', tensile strength T', roughness R', contact angle C', plane resistivity E', and air permeability P'. X'=[H', T', R', C', E', P'] (2) At this time, the GDL aging decay width λ must pass through the following path: It further includes, Mounting the aforementioned gas diffusion layer to be predicted on multiple test stacks of equivalent output and size is, Let N be the stack number. N=[N 1 , N 2 , N 3 ...] (4) Here, 1, 2, and 3 further represent the numbers of the first, second, and third test stacks, Performing a cyclic test of durability aging and decay on each test stack according to the set driving mode of the actual vehicle, and recording the cycle test time for each test stack, Let T be the time taken for each stack cycle test. T=[T 1 , T 2 , T 3 ...] (5) It further includes, After the aforementioned cycle test is completed, the test stacks after multiple aging and decay periods with different cycle test times are disassembled one by one, the gas diffusion layer to be predicted is extracted, and a performance test is performed, and at this time the first average degradation performance of the gas diffusion layer to be predicted is obtained. The first average degradation performance XN of the gas diffusion layer to be predicted after cycle-stack durability of multiple test stacks is as follows: Here, 1, 2, and 3 further represent the numbers of the first, second, and third test stacks, Based on the second average degradation performance and initial performance of the gas diffusion layer to be predicted, the aging decay width of the gas diffusion layer to be predicted after cycle-mounted durability of each test stack is calculated and obtained as follows: The aging decay range of GDL after cycle-based endurance of multiple test stacks is as follows: Here, 1, 2, and 3 further represent the numbers of the first, second, and third test stacks, The method is, The second mean degradation performance XL of the predicted gas diffusion layer after the offline durability test is as follows: At the same time, the aging decay range of the gas diffusion layer to be predicted, obtained by the offline durability test, is as follows: This further includes, Based on the aging decay width, first average performance of degradation, second average performance of degradation, and initial performance of the gas diffusion layer to be predicted, the set of aging decay acceleration coefficients for offline endurance tests and online cycled bench endurance tests is calculated and obtained as follows: The set of aging decay acceleration coefficients θ obtained by dividing the difference between each variable in equation (8) and the initial performance by the difference between each variable in equation (8) and the initial performance is as follows: This further includes, The process of adding the indices of each term in the aforementioned set of aging decay acceleration coefficients to their corresponding values, calculating the average value to obtain the final combination of acceleration coefficients, and then performing an average of these additions to obtain the final acceleration coefficient is as follows: Each of the pending indicators is added to the corresponding value to calculate the average and obtain the final combination of acceleration coefficients, then the summation average is performed to obtain the final acceleration coefficient δ. Equation (11) further includes the setting that all indicators are sensitive to durability aging decay, and that t represents the offline test time. A method for predicting the aging decay of a gas diffusion layer based on a combination of offline and online methods, characterized by the features described above.

Description

This application relates to the technology of fuel cells, and more particularly to a method for predicting the aging and decay of a gas diffusion layer based on a combination of offline and online processes. Durability is a crucial core factor influencing the commercialization of fuel cells. The ability of fuel cells to support stable operation over extended periods is a prerequisite for building consumer confidence and achieving large-scale load-bearing operations. Currently, fuel cells are in a stage of continuous technological maturity, and their durability is directly influenced by the material performance of critical assembly components. While product durability can be determined through actual vehicle operation or bench cycle testing, both methods incur high time and economic costs for completing selection and suitability testing of critical components. Offline single-material testing is lower in cost and more efficient, but it fails to reflect the characteristics of operating conditions in actual vehicle operation or bench cycle testing, resulting in a separation of offline and online performance. To achieve the above objective, this application provides the following solutions. According to a first aspect of the present invention, the present invention is a method for predicting the aging decay of a gas diffusion layer based on a combination of offline and online: To obtain the gas diffusion layer to be predicted and the initial performance of the gas diffusion layer to be predicted, The gas diffusion layer to be predicted is mounted on multiple test stacks of equivalent output and size, Each test stack undergoes a cyclic test of durability aging and decay according to the set driving mode of the actual vehicle, and the cycle test time for each test stack is recorded. After the cycle test is completed, the test stacks after multiple aging and decay periods with different cycle test times are disassembled one by one, the gas diffusion layer to be predicted is extracted, and a performance test is conducted to obtain the first average degradation performance of the gas diffusion layer to be predicted at that time. An offline durability test is performed on the gas diffusion layer to be predicted, the offline durability test time is recorded, and after completion, the second average degradation performance of the gas diffusion layer to be predicted after the adoption of the offline durability test is collected. Based on the second average degradation performance and initial performance of the gas diffusion layer to be predicted, the aging decay width of the gas diffusion layer to be predicted after cycle-mounted durability of each test stack is calculated and obtained. Based on the aging decay width, first average performance of degradation, second average performance of degradation, and initial performance of the gas diffusion layer to be predicted, a set of aging decay acceleration coefficients for offline durability tests and online cycled bench durability tests is calculated and obtained. This includes adding the indices of each term in the set of aging decay acceleration coefficients to their corresponding values, calculating the average value to obtain the final combination of acceleration coefficients, and then performing an average of these additions to obtain the final acceleration coefficients. Furthermore, obtaining the gas diffusion layer to be predicted and the initial performance of the gas diffusion layer to be predicted is The measurement indicators for the gas diffusion layer to be predicted include thickness H, tensile strength T, roughness R, contact angle C, planar resistivity E, and air permeability P. If the initial performance of the gas diffusion layer to be predicted is X, X=[H, T, R, C, E, P] (1) If we set the performance of the gas diffusion layer to be predicted after aging and decay as X', X'=[H', T', R', C', E', P'] (2) At this time, the GDL aging decay width λ must pass through the following path: Includes. Furthermore, mounting the gas diffusion layer to be predicted on multiple test stacks of equivalent output and size is possible. Let N be the stack number. N=[N 1 , N 2 , N 3 ...] (4) Here, 1, 2, and 3 represent the numbers of the first, second, and third test stacks, respectively. Furthermore, performing a cycle test of durability aging and decay on each test stack according to the set driving mode of the actual vehicle, and recording the cycle test time for each test stack, Let T be the time taken for each stack cycle test. T=[T 1 , T 2 , T 3 ...] (5) Includes. Furthermore, after the cycle test is completed, the test stacks after multiple aging and decay periods with different cycle test times are disassembled one by one, the gas diffusion layer to be predicted is extracted, and a performance test is performed, and at this time the first average degradation performance of the gas diffusion layer to be predicted is obtained. The first average degradation performance X