The next generation of Autonomous Underwater Vehicles (AUV) can impact our observation of the world. The flight simulation and full-envelope hydrodynamics modeling can improve the performance of AUVs in terms of control, navigation and positioning. In order to achieve agile maneuverability, a more accurate database of full-envelope hydrodynamic coefficients is supposed to be generated. Two semi-empirical methods, Jorgensen and DATCOM, and two numerical method, Computational Fluid Dynamics (CFD) and XFLR5 are used to push the boundaries of hydrodynamic coefficients: lift, drag and moment coefficients for flight-style AUVs at the Swedish Maritime Robotics Center (SMaRC). A comparison of different approaches and tools, and an analysis of the most appropriate approaches for different regions of a defined maneuver has been conducted in this thesis. A data confidence level was proposed as a way to estimate the accuracy of the data and a structured database was built in terms of data confidence level. Different components of the AUV such as the hull body and wings were analyzed separately. The new database is input to a 3DOF Simulink model and the 6DOF SMaRC hydrobatics simulator for flight dynamics simulations. Simulations show that the new database has a good applicability.