A combination of experimental and model based analysis was performed to investigate calendering impacts on the performance of lithium-ion-batteries. When discharging, not only geometric parameters, such as electrode thicknesses and porosities are affecting performance. Calendering also impacts on other parameters, such as the effective ionic conductivity within the electrolyte, the effective electronic conductivity of solid active material and the effective solid-liquid interfacial area. The simulation supported method is shown to complement experimental analysis to understand correlations between calendering and these parameters; it enables to identify cell internal parameters which are hard to measure and to analyze how the lithium transport is affected. In experiments, cells containing non-calendered cathodes performed significantly worse than ones with 22%-calendered cathodes. Simulation indicated that this losses consist mainly of a deterioration of effective electronic conductivity leading to overpotentials close to the separator. Minor contributions to the losses in non-calendered cathodes caused by the geometric compaction and a reduction of effective solid-liquid interfacial area were found as well, whereas the impact of effective ionic conductivity turned out to be only insignificantly small.

Cell performance of lithium-ion-batteries (LIB) can be tailored to particular hybrid or full electric vehicle applications by targeted adjustment of manufacturing parameters. Furthermore there is a large number of cathode material compositions which can be used. Knowing the correlations between these parameters, electrode structures and cell performance is important to reach the high requirements posed by electromobility. Within this study, impacts of essential manufacturing parameters, being active material mass loading, calendering stress load and carbon black content on the cell performance were investigated for two different, promising cathode materials. For NMC and LMO, the respectively highest calendering stress load and carbon black content yielded the best performance as losses due to poor electronic conductivity were reduced. The active material mass loading rather influenced the ratio between specific energy and specific power. Finally two optimally performing parameter configurations could be identified which were, depending on the required application: NMC with high mass loading and LMO with medium mass loading; in both cases the highest calendering load and carbon black content were applied. An analysis of statistical reproducibility dependent on various parameter configurations was carried out as well. A significant improvement of reproducibility could be achieved by increase of calendering stress load.

The high quality demands of batteries for electric vehicles require powerful tools for error detection in cell manufacturing. Furthermore, cell diagnostics is a serious challenge because performance limitations occur on atomic scale and as batteries are closed systems physical issues can hardly be detected only with the aid of experimental methods. Physico-chemical models enable to detect up to seven various mechanisms of limitations but experimental parameterization is extensive. Therefore, in this study a fast mathematical parameterization approach was used to simulate and diagnose cells with various manufacturing parameter configurations. Limitation mechanisms are shown in correlation with impacts by calendering, electrode thickness, carbon black recipe and cathode active material. Depending on the adjusted production parameters, they vary between low electronic conductivity, overpotentials due to reduction of electrochemically active solid-liquid interfacial area and low ionic conductivity. Furthermore it is shown that characteristic indicators for the particular limitation mechanisms can be observed in discharge curves at various CRates. Finally, a statistical analysis demonstrates how parameter identification can be performed computationally as a side product from reparameterization.