Figure 1 - uploaded by Andreas Wiegmann
Content may be subject to copyright.
3-D and 2-D visualizations of the generated microstructure based on the properties of a SGL10BA carbon paper. The 2-D image is a visualization looking from the top on the image similar to a SEM.
Source publication
In the current work, we present a comprehensive modeling framework to predict the effective gas diffusivity, as a function of liquid water saturation, based on realistic 3-D microstructures of the uncompressed as well as compressed gas diffusion layer (GDL). The presented approach combines the generation of a virtual microscopic GDL and different p...
Contexts in source publication
Context 1
... i.e. the fibers are allowed to overlap • Due to the fabrication process, the fiber system is macroscopically homogeneous and isotropic in the material pane, defined as xy-plane Details about our stochastic model can be found in (2). The reconstructed microstructure of t a SGL10BA, which has been used for further simulations, is shown in Fig. 1. 3.04mm ...
Context 2
... with our reduced model, it is difficult to find a relation between the compression ratio and the external load. Nevertheless, due to the high porosity of the GDL material, our approach leads to a reliable 3-D morphology of the non-woven structures under compression as one can see from Fig. 2. , 3 (1) 1069-1075 (2006) ) unless CC License in place (see abstract). ecsdl.org/site/terms_use ...
Context 3
... the FM model can be used for the uncompressed and the compressed microstructure. Three of the determined capillary pressure curves are shown in Fig. 4. , 3 (1) 1069-1075 (2006 Fig. 4 Capillary pressure-saturation curves for the uncompressed and two different compression ratios. The underlying two-phase distribution corresponds to several drainage and imbibition cycles. ...
Context 4
... our case, we determined a value of 0.74 for the effective diffusion coefficient of the uncompressed sample which is entirely filled by a gas phase. Since the porosity is 0.88, this corresponds to a tortuosity of 1.09 which is still close to one as expected from the 3-D microstructure shown in Fig. 1. , 3 (1) 1069-1075 (2006) ) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 211.71.30.97 Downloaded on 2014-10-13 to IP Using the previous steps of our approach, we study the dependency of the diffusion coefficient on the non-wetting phase saturation and the ...
Context 5
... our case, we determined a value of 0.74 for the effective diffusion coefficient of the uncompressed sample which is entirely filled by a gas phase. Since the porosity is 0.88, this corresponds to a tortuosity of 1.09 which is still close to one as expected from the 3-D microstructure shown in Fig. 1. , 3 (1) 1069-1075 (2006) ) unless CC License in place (see abstract). ecsdl.org/site/terms_use ...
Citations
... 8 Saturation can also influence the tortuosity and reactor resistances as unsaturated pores may increase the length of the ionic pathway, and material stability as reactant-starved surfaces can perform undesirable reactions such as carbon corrosion. 9,10 Irrespective of the electrode geometry or cell design, a pre-requisite for achieving high available surface area is through adequate infiltration of the electrode with the electrolyte phase. ...
Carbon-based porous electrodes are commonly employed in electrochemical technologies as they provide a high surface area for reactions, an open structure for fluid transport, and enable compact reactor architectures. In electrochemical cells that sustain liquid electrolytes (e.g., redox flow batteries, CO2 electrolyzers, capacitive deionization), the nature of the interaction between the three phases - solid, liquid and gas - determines the accessible surface area for reactions, which fundamentally determines device performance. Thus, it is critical to understand the correlation between the electrolyte infiltration in the porous electrode and the resulting accessible surface area in realistic reactor architectures. To tackle this question, here we simultaneously perform neutron radiography with electrochemical measurements to correlate macroscopic electrode saturation/wetting with accessible surface area. We find that for untreated electrodes featuring neutral wettability with water, the electrode area remains underutilized even at elevated flow rates, both for interdigitated and parallel flow fields. Conversely, increasing the electrode hydrophilicity results in an order-of-magnitude increase in accessible surface area at comparable electrode saturation, and is less influenced by the electrolyte flow rate. Ultimately, we reveal useful correlations between reactor architectures and electrode utilization and provide a method that is broadly applicable to flow electrochemical reactors.
... As such, diffusion is restricted to a certain subdomain of the full pore space. In numerical studies [22] as well as in experimental studies [13], a sample's relative diffusion is reported to vary significantly with saturation. However, in general, structure and connectivity of the pore space also influence relative diffusion. ...
In the past several years, convolutional neural networks (CNNs) have proven their capability to predict characteristic quantities in porous media research directly from pore-space geometries. Due to the frequently observed significant reduction in computation time in comparison to classical computational methods, bulk parameter prediction via CNNs is especially compelling, e.g. for effective diffusion. While the current literature is mainly focused on fully saturated porous media, the partially saturated case is also of high interest. Due to the qualitatively different and more complex geometries of the domain available for diffusive transport present in this case, standard CNNs tend to lose robustness and accuracy with lower saturation rates. In this paper, we demonstrate the ability of CNNs to perform predictions of relative diffusion directly from full pore-space geometries. As such, our CNN conveniently fuses diffusion prediction and a well-established morphological model which describes phase distributions in partially saturated porous media.
... These studies typically involve randomly generating GDL structures via a stochastic model. [9][10][11][12][13][14][15] The resulting structure is then subject to a detailed numerical analysis to determine its transport behavior. Alternatively, there have been a number of imaging studies performed which use X-ray tomography. ...
... 10) diffusivities. 1.58 [9] D ρρ m = D 0 (1 − α I ) 3.71 [10] Figure 3b shows this comparison for varying porosity of Toray TGP-H-060 carbon paper. ...
The versatility of a recently developed analytical effective medium theory (EMT) is tested by mimicking the structure of gas diffusion layer materials. The EMT was derived in a previous study [T. D. Myles, A. A. Peracchio, and W. K. S. Chiu, J. Appl. Phys., 115, 203503 (2014); T. D.Myles, A. A. Peracchio, andW. K. S. Chiu, J. Appl. Phys., 117, 025101 (2015)] and is based on the Bruggeman unsymmetrical theory. This EMT calculates the effective transport property of a mixture consisting of multiple types of inclusion particles, which are spheroid shaped and have varying orientation anisotropy, embedded in a continuous matrix. Each inclusion type may have a unique bulk transport property, orientation anisotropy, and shape. As a starting point, the theory was first tested against a commonly cited numerical result from the literature for effective diffusivity through fibrous structures. From there the EMT was modified to account for web-like structures that develop in Toray carbon paper materials. Finally, the in-plane thermal conductivity of the carbon paper was predicted. The results show good agreement to available numerical and experimental data in the literature despite the relatively simple expressions and implementation. However, one issue which was highlighted in this study was the lack of accounting for percolation phenomena in the developed EMT.
... Wang et al. further presented detailed DNS to disclose the transport phenomena of mass, reactant, electron, and heat occurring inside the GDL, see Figs. 27 and 28 [174]. Wang and co-workers [175][176][177] applied the LBM (Lattice Boldtzman method) to study the meso-scale transport of liquid water, based on detailed GDL structure either from stochastic modeling or experimental imaging (e.g. X-ray micro-CT). ...
Polymer electrolyte membrane (PEM) fuel cells, which convert the chemical energy stored in hydrogen fuel directly and efficiently to electrical energy with water as the only byproduct, have the potential to reduce our energy use, pollutant emissions, and dependence on fossil fuels. Great deal of efforts has been made in the past, particularly during the last couple of decades or so, to advance the PEM fuel cell technology and fundamental research. Factors such as durability and cost still remain as the major barriers to fuel cell commercialization. In the past two years, more than 35% cost reduction has been achieved in fuel cell fabrication, the current status of $61/kW (2009) for transportation fuel cell is still over 50% higher than the target of the US Department of Energy (DOE), i.e. $30/kW by 2015, in order to compete with the conventional technology of internal-combustion engines. In addition, a lifetime of ∼2500 h (for transportation PEM fuel cells) was achieved in 2009, yet still needs to be doubled to meet the DOE’s target, i.e. 5000 h. Breakthroughs are urgently needed to overcome these barriers. In this regard, fundamental studies play an important and indeed critical role. Issues such as water and heat management, and new material development remain the focus of fuel-cell performance improvement and cost reduction. Previous reviews mostly focus on one aspect, either a specific fuel cell application or a particular area of fuel cell research. The objective of this review is three folds: (1) to present the latest status of PEM fuel cell technology development and applications in the transportation, stationary, and portable/micro power generation sectors through an overview of the state-of-the-art and most recent technical progress; (2) to describe the need for fundamental research in this field and fill the gap of addressing the role of fundamental research in fuel cell technology; and (3) to outline major challenges in fuel cell technology development and the needs for fundamental research for the near future and prior to fuel cell commercialization.
... The direct simulation is a powerful numerical tool for studying the transport in porous media. Some also named it the direct numerical simulation (DNS)232425. To distinguish with the terms in the computational fluid dynamics (CFD), e.g. in the turbulent simulation, we follow Kaviany [26,27] and choose the term of direct simulation for this numerical technique. ...
This paper combines the stochastic-model-based reconstruction of the gas diffusion layer (GDL) of polymer electrolyte fuel cells (PEFCs) and direct simulation to investigate the pore-level transport within GDLs. The carbon-paper-based GDL is modeled as a stack of thin sections with each section described by planar two-dimensional random line tessellations which are further dilated to three dimensions. The reconstruction is based on given GDL data provided by scanning electron microscopy (SEM) images. With the constructed GDL, we further introduce the direct simulation of the coupled transport processes inside the GDL. The simulation considers the gas flow and species transport in the void space, electronic current conduction in the solid, and heat transfer in both phases. Results indicate a remarkable distinction in tortuosities of gas diffusion passage and solid matrix across the GDL with the former ∼1.2 and the latter ∼13.8. This difference arises from the synthetic microstructure of GDL, i.e. the lateral alignment nature of the thin carbon fiber, allowing the solid-phase transport to occur mostly in lateral direction. Extensive discussion on the tortuosity is also presented. The numerical tool can be applied to investigate the impact of the GDL microstructure on pore-level transport and scrutinize the macroscopic approach vastly adopted in current fuel cell modeling.
Water management in proton-exchange membrane fuel cells (PEFCs) has a large impact on the performance of the device, as liquid water affects the transport properties of the gas diffusion layer (GDL). In this study, we develop an ensemble-based model of the liquid water distribution inside the GDL. Based on a water injection experiment, the wet structure of the porous medium is inspected via X-ray tomographic microscopy and, after an image segmentation process, a voxel-based meshing of the fiber, air, and water domains is obtained. Starting from the obtained dry fiber structure, a Metropolis-Hastings Monte Carlo algorithm is used to obtain the equilibrium distribution of liquid water that minimizes the surface free energy of the ensemble. The different water distributions from the Monte Carlo (MC) simulation and water injection experiment are identified as solution for different physical mechanisms both of which are present in a running fuel cell. The wet structure is then used to calculate saturation-dependent effective transport properties using the software GEODICT. Thereby, a strong influence of the saturation gradient on the macrohomogeneous transport properties is found.
This paper presents a methodology for modeling microstructures of fibrous porous media with curved fibers. The developed methodology utilizes implicit periodic surface model coupled with the genetic algorithm (GA) optimization to construct the porous microstructures. The fibers profile is represented by the periodic implicit surfaces and their orientation and curvature are determined by GA optimization. To reconstruct the microstructures with higher resemblance to the actual porous media GA is utilized to minimize the fibers stored strain energy and their intersection volumes. Coupling the image processing techniques to the geometry construction procedure the morphological and transport properties of the constructed microstructures are also determined. To verify the feasibility and the accuracy of the proposed methodology the microstructure of Freudenberg H2315 GDL is constructed and characterized. The presented methodology enables a parametric design approach. Thus, the effects of the microstructure's properties e.g., fibers diameter, fibers orientation and porosity of the porous structure on the transport properties of the fibrous media are investigated.
The gas diffusion layer (GDL) is one of the key components in a polymer electrolyte membrane (PEM) fuel cell. Generally it is a carbon-based fibrous medium that allows for the transport of electrons through the fibers and distributes the reactants through the void space to the catalyst layer in a PEM fuel cell. In the present work, a microstructure study of reactant transport is carried out by reconstructing the typical fibrous microstructure of the GDL and investigating the transport characteristics of the porous medium using computational fluid dynamics (CFD) simulations. The results confirm the applicability of Darcy’s law formulation for permeability determination and Bruggemann correction for calculation of effective diffusivity for typical conditions encountered in PEM fuel cells. Macroscopic material properties such as through-plane and in-plane permeabilities and effective diffusion coefficient are determined and compared against experimental values reported in the literature.
