Proton exchange membrane fuel cells have been identified as suitable for vehicle implementation. As for all fuel cell types, thermal management is a critical issue: the redox reactions taking place in the membrane electrode assembly generate water; a humid membrane is able to enhance transport phenomena, but too much water could flood the pathways of reactant gases. Recently, interest has grown with respect to the properties of the gas diffusion layer of membranes: its fibres pattern can critically affect water removal rates. Computation fluid dynamics represents a powerful tool that can be used for understanding the key design factors of fuel cell components to improve the overall performance. Specifically, in the case of the diffusion layer, 3D simulation can investigate droplet dynamics and define the optimized surface structure for water management. In this paper, optical data and numerical results were combined to characterize the behaviour (oscillation, detachment, etc) of a couple of droplets on the woven gas diffusion layer with a structure retrieved by using an approach typical of the textile industry. High-spatial resolution imaging allowed the evaluation of warp/weft size; these parameters were used for building the 3D mesh for simulations based on the Volume of fluid method. The dynamics of droplets, with diameters ranging from 200 to 600 μm, were investigated under the effect of a constant 10 m/s airflow for different gas diffusion layer geometries. In addition, the cases of the droplet pairs, deposited at different reciprocal distances, were studied by using the same methodological approach. Results demonstrated the impact of layer structure on the results obtained through simulations; moreover, an optimised design can contribute to control water removal and minimise flooding effects.

Nowadays reducing green-house gas emissions and pushing the fossil fuel savings in the field of light-duty vehicles is compulsory to slow down climate change. To this aim, the use of new combustion modes and dilution strategies to increase the stability of operations rich in diluent is an effective technique to reduce combustion temperatures and heat losses in throttled operations. Since the combustion behavior in those solutions highly differs from that of typical market systems, fundamental analyses in optical engines are mandatory in order to gain a deep understanding of those and to tune new models for improving the mutual support between experiments and simulations. However, it is known that optical accessible engines suffer from significant blow-by collateral flow due to the installation of the optical measure line. Thus, a reliable custom blow-by model capable of being integrated with both mono-dimensional and three-dimensional simulations was developed and validated against experimental data. The model can work for two different configurations: (a) stand-alone, aiming at providing macroscopic data on the ignitable mixture mass loss/recover through the piston rings; (b) combined, in which it is integrated in CFD engine simulations for the local analysis of likely collateral heat release induced by blow-by. Furthermore, once the model was validated, the effect of the engine speed and charge dilution on the blow-by phenomenon in the optical engine were simulated and discussed in the stand-alone mode.