In this master´s thesis special flow phenomena of the powder injection molding process were experimentally and numerically investigated since the physical backgrounds behind the flow behavior of the used materials (feedstocks – highly filled plastics) cannot be fully described yet. Consequently, nowadays there are still major discrepancies between the simulation of the injection molding process with feedstocks and the practical experiment. Feedstocks exhibit a higher thermal conductivity, a higher heat transfer and a lower specific heat compared to unfilled plastics and it was hypothesized, that a certain threshold temperature exists, where the material suddenly changes from fluid to a solid behavior. Whereas the material is expected to flow in a channel at high temperatures like unfilled thermoplastics, below this threshold temperature the material will only be pushed through the channels like a solid. The material close to the wall will reach this threshold temperature very quickly and form a solid case, which will slip at the wall. With experimental filling studies with two different cavities and three different materials, the flow behavior was visualized at different mold temperatures and injection rates and the accordance with the hypothesis was analyzed. This hypothesis was rebutted by the experiments. Although at low mold temperatures the material was pushed through the cavity as solid, no threshold temperature could be determined since there was a continuous transition of the flow behavior at higher mold temperatures. A typical fountain flow of the material could not be achieved below the melting temperature and special flow phenomena appeared. There was always a preceding material area at the melt front and the material tended to keep its shape even at changes of the cross section of the flow channel. The measured injection pressures showed a linear correlation with the mold temperature and no sudden changes due to an existing threshold temperature could be observed. The simulation of the experiments predicted much lower pressures (average deviation of 69 % to the real pressure) and showed a completely different flow behavior comparable to unfilled plastics (like polypropylene) with standard settings in the software. In fact, none of the observed flow phenomena was reproduced by the simulation, which highlights the importance of understanding the physical processes.