A recent advance in the field of turbulence modeling is the development of hybrid turbulence models. These models, such as Detached Eddy Simulation (DES) [Spalart et al, 1997] merge Reynolds-Averaged Navier-Stokes (RANS) models with Large Eddy Simulation (LES) and/or Direct Numerical Simulation (DNS). With an increasing number of these proposed models, there is an increasing need to determine the capabilities and limitations of these new models. We will present an ongoing approach to evaluate these new turbulence models through a combination of careful validation against experimental results and comparison to DNS results. Using the common platform LESTool, a fast, adaptable CFD code designed to use RANS, DES, LES, or DNS, we can make direct comparisons between all these turbulence models. A particular focus of this validation program is whether the computational simulations vary smoothly between the LES/DNS results (for extremely fine grids) and the RANS results (for coarse grids) as the grid density and numerical schemes are varied. Numerical schemes under investigation include high-order upwind schemes, the hybrid central/upwind scheme of Travin et al [2000] and Strelets [2001], and the ENO-Pade scheme developed at the University of Kentucky [Wang et al, 2000].