This work presents a computational strategy for the numerical simulation of Additive Manufacturing (AM), focusing specifically on the Direct Energy Deposition (DED) process applied to casting structures used as substrates. The proposed solution involves the automatic generation of a finite element (FE) mesh to discretize the casting component by embedding the corresponding CAD file into the background mesh. An octree-based mesh refinement approach is utilized to ensure an accurate definition of the component boundaries. The AM process is then simulated using a growing computational domain, dynamically generated based on the defined scanning path (e.g., G-code format). The coupled thermo-mechanical simulation evaluates both the temperature field induced by the DED process and the resulting residual stresses. To identify the optimal process parameters, the strategy adjusts the energy density (defined by the power supply and the scanning speed) in response to the current substrate temperature. Additionally, the optimal dwell time between successive layers is calculated to minimize alterations to the original microstructure at the interface between the casting and the AM build. The same software framework is extended to support repair operations using the DED process.