The development of IC engines is a complex process where 0D/1D-simulation tools became more important in the past few years. Different designs can be investigated in very early stages of the development process without the expensive buildup of prototypes and it is possible to get reliable results with passable effort. The quality of the overall simulation results depends on the quality of the sub-models. Simulation of the combustion process in natural-gas SI engines relies on predictive models for burn rates and knock. Existing knock models for gasoline fuels are based on a time-integrated ignition delay, using a fitted Arrhenius equation. Within a research project an enhanced knock-model approach for methane based fuels was developed. Chemical kinetics models were used to calculate the auto-ignition times for various temperatures, pressures and air-fuel-ratios (AFR). Furthermore, auto-ignition has been investigated for binary CNG substitutes by admixing different amounts of ethane, propane, butane and hydrogen to methane as a main component. Based on these results, a kinetic fit was derived and the auto-ignition model was validated with test bench measurements. Considering single cycle analysis of the measurement data, the enhanced model is able to simulate the onset of auto-ignition in the unburned zone. The knock-model approach is embedded in a quasi-dimensional burn rate model to describe the combustion process in natural gas powered SI engines. Therefore, the laminar flame speeds of different binary CNG substitutes were calculated in Cantera and approximated with a correlation. Besides the influence of the fuel composition, the effect of temperature, pressure, AFR and EGR-rate on the laminar flame has been investigated. When focusing on different fuel compositions and their effect on combustion and auto-ignition in natural gas powered SI engines, the improved models provide good and predictive results for 0D/1D-simulations.