Computational modeling provides significant benefits in assessing the helmet performance and identifying promising helmet designs. We develop multi-fidelity computational tools, representative virtual human head and helmet system models to help the design of next generation combat helmet with improved protection against blunt impact. By integrating the fast-running articulated human with personal protective equipment (PPE) biodynamics model with the high-fidelity human head with combat helmet finite element (FE) model, the multi-fidelity approach can be used to efficiently investigate impact-related traumatic brain injury (TBI) in the real-world scenario. The FE model is used to capture the dynamics of the composite helmet shell, foam pad suspension, retention strap and head while the biodynamics model provides the proper kinematics and boundary conditions for the FE model. An orthotropic elasto-plastic material with damage model is employed for the helmet shell. Enhanced tetrahedral elements are used to model the nearly-incompressible tissues. The head with helmet and without helmet under a severe impact due to a fall caused by blast loading are simulated and compared. The resulting biomechanical responses of head acceleration, shear stresses and strains in brain and mechanical injury criterion as well as helmet energy absorption are used to characterize the performance of helmet system.