Although recent developments in adaptivity and rezoning have enhanced the ease and use of the finite element method, the question remains: how good are the resulting simulations? In the case of highly nonlinear behavior, it appears that this question can only be answered by properly designed experiments. One such experiment involves the nonlinear behavior (material, geometry and boundary conditions) associated with a precision closed-die (PCD) forging. This paper presents some PCD data that could not be modeled by one particular FE code-simulation. Hopefully, these data can be used for the quantitative error-assessment of other codes for similar nonlinear simulations. Some possible code-simulation improvements are discussed and recommendations for future experiments are made.