Rapid development in computing power in past five decades along with the development and progress in building blocks of Computational Fluid Dynamics (CFD) technology has made CFD an indispensable tool for modern engineering analysis and design of fluid-based products and systems. For CFD analysis, Reynolds-Averaged Navier-Stokes (RANS) equations are currently the most widely used fluid equations in the industry. RANS methods require modeling of turbulence effect (i.e. turbulence modeling) based on empirical relations and therefore often produce low accuracy results for many flows. In recent years, the Large Eddy Simulation (LES) approach has been developed which has shown promise of achieving higher accuracy, however it is computationally very intensive and therefore has remained limited to computing relatively simple flows from low to moderate Reynolds numbers. As a result, a hybrid technique called Detached Eddy Simulation (DES) has been proposed in recent years. This technique has shown improved accuracy and computational efficiency for solution of wide variety of complex turbulent flows. The goal of this dissertation has been to develop a DES model based on a recently proposed very promising RANS model, known as the 'Wray-Agarwal (WA)' model and the LES. Decaying Isotropic Turbulence (DIT) case is computed to determine the coefficient in the DES model by matching its energy spectrum with the Kolmogorov spectrum. The new WA-DES model (DES model based on WA model) is applied to compute a wide variety of wall bounded separated flows to assess it accuracy and computational efficiency compared to the widely used RANS turbulence models in the industry, namely the Spalart-Allmaras (SA) and SST k-[omega] models. Improved Delayed-Detached Eddy Simulation (IDDES) and Elliptic Blending are also considered as further refinements of WA model to improve its accuracy.