Sheet metal forming is one of the most common forming processes used in the industry, especially in the automotive industry. It becomes a common sense, that by increasing the accurate simulation of sheet metal forming, the industry can save dramatic cost in trial-and-error process when designing the sheet metal forming tools. In past decades, considerable studies have been done in the field of numerical analysis of metal forming processes, particularly in springback prediction. Significant progresses have been made, but the accuracy of simulation results still needs to improve. One reason is that, in the typical sheet metal forming analysis the tools are considered as absolute un-deformable rigid bodies. The deformation of the tools, which happens in the real production, is not taken into account. Another reason may be that the classic simulation considers the friction coefficient between the tools and blank as constant. However, the actual friction condition depends on a number of parameters. The objective of this thesis work is trying to investigate how it will affect the springback prediction results when either the tool deformation or more complex friction conditions are considered. The purpose is nothing related with the precise simulation about the true problem or how accuracy the simulation results show compared to the experiment results. The work is only to give an emphasis hint that how the FE-model and friction mode chosen affects the springback results when doing a numerical analysis. A simple model called flex rail is used for sheet metal forming simulations with three different friction models. A comparison between the results clearly shows a difference when advanced friction models are applied. A 3D elastic solid model is created to compare the result with rigid model. The results show the difference when deformation of the tools is taken into account. Finally, an actual case with tools from the industry is investigated. The tools are from SAAB Cars Body Components. This case is to investigate the possibility and necessity of applying the advanced friction model and elastic tools when a complex real industry problem is faced. Further study is needed to do with comparison experimental data to verify the accuracy when these models are used.