Study On Design And Optimization Of Novel Flow Field Structures For Proton Exchange Membrane Fuel Cells

The proton exchange membrane fuel cell(PEMFC)is a device that directly converts the chemical energy of fuel into electrical energy.It has received more and more attention because of its advantages of low operating temperature,fast start-up,high conversion efficiency,no noise and vibration,and wide range of applications.As one of the key components of PEMFC,the bipolar plate plays the role of mechanical support,electrical conductivity and heat sinking.The configuration of the flow channel on the bipolar plate affects the reaction gas supply capacity,distribution uniformity,and liquid water removal capacity,and it plays a critical role in the improvement of the fuel cell performance and long-term stable operation.In order to improve the performance of PEMFC,this paper integrates numerical simulation,surrogate models,and optimization algorithms to implement a series of optimization.The main contents are drawn as follows.First,a three-dimensional(3-D),multiphase,steady-state mathematical model of the proton exchange membrane fuel cell is established,which includes the following equations:the basic conservation equation,the electrochemical equation,and the water formation&transport equation.The model boundary conditions and model assumptions are also defined.The optimization framework formed by coupling the surrogate model with the genetic algorithm is applied to optimize the channel cross section of the three-dimensional straight channel PEMFC.The topline/baseline bottom width of the straight channel are chosen as the optimization variables,and the net output power of the PEMFC is used as the objective function,and the trained Kriging surrogate model is applied instead of the computational fluid dynamics model of the PEMFC.Then,the genetic algorithm is applied to perform global optimization and the geometry parameters of the optimization model are obtained.The flow channel structure of the optimization model is a wide and short trapezoidal structure,and the topline/baseline bottom widths of the optimization model are 0.9993 mm and 1.5998 mm,respectively.The maximum net output power of the optimization model is increased by 25.94%compared with the base model.In addition,the oxygen concentration gradient of the optimal model is lower than that of the basic model along the flow direction of the reaction gas and the perpendicular direction to the flow,and the oxygen distribution of the optimization model is more uniform.And the optimization model has a superior water removal capacity.Secondly,the bionics theory is applied to the PEMFC flow field design,and a new leaf vein bionic flow field(LVFF)is designed.The effect of the number of flow field branch channels on the fuel cell performance is investigated by numerical simulation.With the increase of the number of branch channels,the output power density of PEMFC increases and then decreases,and the oxygen distribution inside the PEMFC is more uniform.However,the reaction gas flow rate is too low and oxygen supply is insufficient in the end of the main channel of LVFF with more branched flow channels.Therefore,the streamlined blocks are added to the branched flow channels of the optimal performance LVFF to construct a composite bionic flow field.The reasonable arrangement of the streamline blocks makes the reaction gas distribution more uniform and reduces the risk of gas backflow.In addition,the streamlined blocks promote the oxygen transfer without producing larger parasitic power.And the maximum power density of the CBFF is increased by 8.475%compared with the serpentine flow field.Finally,a novel 3-D fine-mesh flow field for PEMFC was designed.Taking the key structural characteristics of the flow field as variables and the PEMFC net power density and oxygen uniformity index as objective functions,the mapping relationship between the flow field structure and the objective functions was established by applying artificial neural network(ANN)surrogate models.After that,the single-objective optimization(SOO)model and the multi-objective optimization(MOO)model were obtained by applying genetic algorithm and NSGA-II,respectively.The comparison of the two models by numerical simulation shows that although the maximum net power density of the SOO model is slightly higher than that of the MOO model,the overall pressure drop of the flow field of SOO model is significantly higher than that of the MOO model,and the molar concentration of oxygen,oxygen uniformity index and water removal capacity are lower than those of the MOO model.The total performance of the MOO model is better than that of the SOO model.The flow field designed with multiobjective optimization is more conducive to improving the overall performance of the fuel cell and ensuring the long-term stable operation of the cell.