During winter periods, fuel cell systems, particularly for automotive fields of application, could be exposed to extreme climatic conditions. Thereby, repeated cycles between +80°C down to -40°C over short periods of time are possible just as longer soaking times at temperatures far below 0°C with starts from frozen conditions. This could include high numbers of freeze-thaw cycles, depending on the type of application and the corresponding climate zone. Basically, in fuel cells the existence of residual water within the membrane or the gas diffusion layer after shut down and the associated ice formation at temperatures below 0°C represents a problem for membrane-electrode assemblies. Beside this, porous materials contain different pore sizes, so that different water modifications can coexist in a wide temperature range. This results during cooling after freezing of the macroscopic water into redistribution of unfrozen water and continuous growth of ice lenses as well as shrinkage phenomena within the denser pore structure (theory of micro ice lens formation). The removal of residual water from the cell for instance by means of purging, involves also different disadvantages, besides the fact that some water still remains inside the cell, especially in porous media. Here, the observed damage phenomena differ in dependence on the test arrangement and test procedure, as no standardized test method for the estimation of the behaviour of fuel cell systems under cyclic freeze thaw attack exists so far. Therefore the damage mechanisms are not fully understood to date.
Within the actual research project the influence of several boundary test conditions such as cooling/freezing rate, minimum temperature, soaking time as well as cell humidity is analyzed in detail on the damage behaviour of a previously defined PEM standard stack. Main focus is laid on the simulation of the real existing boundary conditions, so that a climatic attack, which is given in practice with the appropriate damage mechanisms, can be simulated in the laboratory. Here, it has to be distinguished between different loading conditions such as cyclic freezing and thawing as well as start up from temperatures below 0°C (cold start). After standardization of the test procedure and test arrangement, damage classes as well as assessment criteria will be defined, in order to estimate the durability of a membrane- electrode assembly with regard to freeze-thaw resistance. In the end a test application will be developed for further uniform testing of PEM at low temperatures.