A holistic model for life time prognosis is most essential for a Li ion battery to be used in an electric vehicle. In particular, it is required for a valid life time prognosis of calendric life time. The prognosis should cover a possible life time of 10 to 15 years . While high mileage can be tested by cycling in a relatively short period of time, this is not possible for the long calendric life time itself. In this thesis, a calendric aging model for NMC vs. graphite Li-ion cells is developed. As a first step, a check-up procedure is developed that can determine usable as well as the mobile Li capacity. The mobile Li capacity is the capacity that can virtually be discharged from a cell with an infinitesimal small current. The holistic aging model is considering the following three aging mechanisms: Anodic side reaction (ASR), cathodic side reaction (CSR), passive anode effect (PAE). The aging trend of the mobile Li capacity is found to be square-root shaped, which has often been reported for ASR. Only due to large internal resistance at end of discharge, the usable capacity, that determines the capacity which can be used in a vehicle application, might not follow this trend. With a half-cell based electric model the different aging trends are made plausible and comparable. Finally, it is shown that the aging at 70% SoC is indeed dominated by ASR. At 100% SoC a significant amount of CSR is found. As a result of CSR, Li is again stored in the cathode. Using the half-cell based model, it is shown that this effect is compensating a part of the loss of mobile Li caused by ASR. Thus, the cell shows less capacity fading at 100% compared to 70% SoC. On the other hand, the internal resistance is strongly increasing during storage at high SoC, likely due to the electrolyte oxidation during CSR. For the parametrization of the aging model, a multi-step approach is chosen. This is used to verify the chosen mathematical descriptions of the aging mechanisms driven by electrode voltage and temperature. Thereby, it is shown that the model can be used to extrapolate from a few test conditions to lower SoC and temperatures.weiterlesen