Rotary draw bending is a widely employed process for manufacturing high-precision metallic tubular components, but performance may be limited by defects such as wall thinning, wrinkling, and cross-section distortion (ovalization). Additionally, spring-back, where elastic recovery occurs after elastoplastic bending, must be accurately predicted to control the process and avoid deviations in bending angle, radius, and final geometry. To this end, and following a comprehensive design of experiments, this work presents a combined numerical-experimental study aimed at characterizing spring-back and elongation in several metallic materials, including SS304, S235, Fe510, Ti99.5, and Al2017A, across distinct tube wall thicknesses and bend radii. Parameterized finite element models were developed to simulate rotary draw bending and subsequent spring-back. Numerical and experimental results showed similar tendencies, where materials with higher elastic moduli exhibit more pronounced elastic recovery. Additionally, the experimental campaign produced a series of bent tube specimens that were digitized using a Hexagon 3D scanning measurement system, enabling direct comparison of the measured geometries and their deviations with the corresponding numerically generated shapes. Furthermore, the AMOB eMOB52 rotary draw bending machine used in the experimental campaign was instrumented to measure its structural stiffness and the reaction forces and deflections at the tool supports. These structural results are then used to refine, calibrate, and validate the corresponding numerical models. After calibration, the ultimate goal is to generate a surrogate machine learning model capable of replicating the parameterized finite element simulations, enabling rapid sensitivity analyses and accelerating process optimization in industrial tube-bending applications. Acknowledgements: This work was funded by the European Regional Development Fund through the project "Smartubending – Curvadoras inteligentes, eficientes e conectadas" (COMPETE2030-FEDER-01480000).