干湿循环条件下膜电极屈曲变形的力学衰减机理分析
Mechanical Degradation Mechanism of Membrane Electrode Assemblies in Buckling Test Under Humidity Cycles

Tomoaki Uchiyama                                                
Manabu Kato                                  
Yoshihiro Ikogi                                  
Toshihiko Yoshida

Membrane electrode assembly （MEA） buckling tests in microscopic clearances under humidity cycles and numerical analyses by finite element method （FEM） were conducted． The NR211 （Dupont， 25－μm thickness， equivalent weight （EW） ＝ 1100） sandwiched between catalyst layers （CLs） was used as the MEA． Based on tensile tests of the NR211 and NR211－CL and FEM simulation of tensile tests， the Young’s modulus and yield point of CL were estimated． While the CL had a higher Young’s modulus than the NR211 in water vapor， the CL indicated a lower Young’s modulus than the NR211 in liquid water at 80 °C． The buckling tests in microscopic diameter of 200 μm in polyimide film were carried out． The heights of bulge in the NR211 and NR211－CL after five humidity cycles were measured with a laser microscope． The height of the NR211－CL was lower than that of the NR211， due to the stiffer CL and the lower swelling ratio of the NR211－CL． Moreover， when the humidity cycles were repeated less than 1000 times， cracks were formed in the CL． The stress－strain behaviors of the NR211－CL buckling test under a humidity cycle were investigated by using the FEM． When the NR211－CL swelled， higher stress was developed at the topside of bulge and topside of bulge round． These portions corresponded to the CL crack－formed portions in the buckling test． When the NR211－CL deswelled， the tensile stress was induced in the entire NR211． The mechanical degradation mechanisms were considered as follows： Firstly， cracks initiate and propagate in the CL when the MEA swells in repeating humidity cycles． Moreover， the tensile stress is induced in the polymer electrolyte membrane （PEM） under deswelling and the CL cracks propagate into the PEM from the CL， which results in pinholes in the PEM．

这篇文章是干湿循环条件下膜电极屈曲变形造成催化层裂缝的原因分析与对策之前的文章，实验条件稍有不同，更注重实验技术和一些细节的描述，我主要了解一下读昨天文章不理解的地方这篇文章中是否有解释。

Fig． 1 Buckling test of NR211 and NR211－CL under humidity
cycles

PI膜的直径不同，干湿循环条件下膜电极屈曲变形造成催化层裂缝的原因分析与对策的直径是300微米。

Fig． 2 Elastoplastic parameters for materials

材料是弹塑性特性。

A data of NR211 was used for simulating the stress and strain．
Even though a slight anisotropy in the NR211 was observed， in this study， for simplicity， material properties are dealt with as isotropic．

From the linear elastic layer estimation calculated， the Young’s modulus of the CL （ECL） was estimated． where E is the Young’s modulus and t is thickness

Poisson’s ratio of the CL was assumed to be 0．25

The GDL was assumed linear－elastic with no swelling． The GDL properties were assumed： Er＝Eh＝1000 MPa，Ez＝20 MPa， and Poisson’s ratio＝0．13

我对泊松比没有概念，但惊讶于这个值怎么来的。我敢打赌你没有阅读过泊松比来源的原文，因为原文的下载量只有15次。引用了

The Mechanical Changes in the MEA of PEM Fuel Cells due to Load Cycling

中的数据

在我看来其中有些值有一定争议，但仿真的似乎能够自圆其说就行，不太考虑绝对值准确与否，但我认为只关心输出的结论不关心输入和对标这其实是个很坏的习惯。

Furthermore， the CL was assumed with no swelling， because CL has a pore in its structure and the pore would absorb the ionomer swelling in the CL

有催化层后复合材料的干湿维度变化有一定变化。

干湿循环条件下膜电极屈曲变形造成催化层裂缝的原因分析与对策中In－plane swelling （εin） was found to be 12．7％， and through－plane swelling （εth） was 24．6％．

干湿循环条件下膜电极屈曲变形造成催化层裂缝的原因分析与对策和本文有引用关系，但数据存在一定差异，而两篇文章对测量方面的问题没有深入探讨。

Fig． 6 Stress－strain curves of NR211 and NR211－CL for （a） 5％ RH， （b） 40％ RH， （c） 80％ RH， （d）100％ RH

数据测量是个吃力不讨好的工作，但及其重要，这个实验是控湿度的高温原位拉伸，没那么好做。

Samples were removed from the PI film after the test， the deformations were observed， and the heights of bulge were measured by a laser microscope （Olympus， Japan） at 25 C， 50％ RH．

是塑性的残余应变。

Fig． 7 Bulge deformations of NR211 and NR211－CL by laser
microscope observations． （a） NR211 after five cycles and （b）
NR211－CL after five cycles．

Fig． 8 Height of bulge after humidity cycles． Error bar indicates
maximum and minimum values．

Fig． 9 Laser microscope observations for bulge deformations of NR211－CL． （a） After five cycles （height of bulge 3．6 um）， （b） after 1000 cycles （height of bulge 15．6 um）， （c） after 4000 cycles （height of bulge 28．8 um）．

Fig． 10 Mises stress distributions in NR211 simulation． Comments indicate portion， dominant stress， and Mises stress value． （a） At 50％ RH （step 1）， （b） at 100％ RH （step 2）， （c） at 50％ RH （step 3）， and （d） at 5％ RH （step 4）．

Fig． 11 Plastic equivalent strain distribution in NR211 simulation
（step 5）

Fig． 12 Mises stress distributions in NR211－CL simulation． Comments indicate portion， dominant stress， and mises stress value． （a） At 50％ RH （step 1）， （b） at 100％ RH （step 2）， （c） at 50％ RH （step 3）， and （d） at 5％ RH （step 4）．

On the contrary， the compressive stress inside the bulge does not affect the CL cracks．

3MPa没到屈服点，不会直接拉坏，应该再做个疲劳测试。

Fig． 13 Plastic equivalent strain distribution in NR211－CL
simulation （step 5）

Fig． 14 Strain behaviors with repeating humidity cycles． （a）
True strain in 0–0．4， （b） true strain in 0–0．02， （c） relative humidity
change．

不太理解黑色线真应变随周期变化逐渐减小的意思。

Conclusions

In the present work， the mechanical degradation mechanism of
the MEA under humidity cycles was considered using the MEA
buckling test and FEM analysis against the buckling test． In the
MEA buckling test under humidity cycles， the MEA bulge grew
with humidity cycles． In addition， CL cracks were formed and

propagated． The stress－strain behaviors in the buckling test with
the NR211 and NR211－CL were analyzed with FEM． The material
properties of elastoplasticity in the NR211 and CL were measured
at several relative humidities． The simulation of the NR211
test showed similar bulge with the experiment， and the simulation
was validated．

From the FEM analysis of the NR211－CL buckling test， stressstrain
behaviors were investigated． Higher Mises stress positions
in the CL corresponded to the CL crack locations in the NR211－
CL buckling test． Namely， biaxial tensile stress was developed at
the topside of the bulge and shear stress was induced at the topside
of the round when the NR211 swelled， resulting in the CL cracks
with repeating humidity cycles． Moreover， the tensile stress was
induced in the entire NR211 under deswelling． The CL cracks can
propagate into the NR211 from the CL， leading to the pinhole in
the NR211． Therefore， the CL crack formation is critical for the
mechanical durability of the MEA． It is our expectation that the
data from these experiments and analyses will provide fundamental
understandings on the mechanical degradation mechanisms of
the MEA．