As a widely used metal additive manufacturing technology, directed energy deposition (DED) provides a cost-effective solution for repairing and remanufacturing high-performance metallic components. Nevertheless, the industrial application of DED is limited by inherent shortcomings, such as excessive geometric distortion (caused by high-gradient bulk residual stress) and poor surface quality (due to the layer-by-layer material addition manner and the presence of incompletely melted feedstocks). To achieve desired surface finishes, post-machining of DED is usually performed. However, due to the thermo-mechanical loads exerted by the cutting tool during material removal processes, additional surface residual stress will be introduced on the DED components. These bulk and surface residual stresses can compromise structural integrity, leading to geometrical/dimensional inaccuracies, shortened fatigue life, and increased susceptibility to corrosion. Guided by three critical and practical research questions—(1) how to efficiently predict the DED-induced bulk residual stress, (2) how to accurately simulate the surface residual stress generated by machining, and (3) how DED-induced bulk stress interacts with machining to dress the final surface stress—my Ph.D. thesis develops physics-based models to understand the formation mechanism of residual stresses in combined DED and machining processes, which is essential for designing optimized process strategies that ensure satisfactory component performance. First, analytical models are proposed to determine thermo-mechanical responses in single-material and functionally graded DED processes. These models employ finite difference methods for predicting temperature cycles, modified Green’s functions to quantify thermal stress, and radial return methods to update plastic stress, which features the incremental material addition as the heat source moves in each deposition layer in DED. The analytical frameworks effectively demonstrate how DED process parameters, deposition strategies, and material gradients influence the bulk residual stress distribution. Second, finite element method (FEM) models are presented to predict machining-induced surface residual stress in (1) conventional wet machining processes (aimed to extend tool life), where tool/workpiece/fluid interactions are dynamically modelled, and (2) high-speed machining processes (designed to enhance productivity), where serrated chip formation imposes time-varying thermo-mechanical loads on machined surfaces. The results quantitatively illustrate the mitigation of surface tensile residual stress due to applying cutting fluids and the fluctuation of surface residual stress in the cutting velocity direction coupled with shear localization in chips. Lastly, the FEM-based machining models are extended by incorporating the initial stress states and microstructure features, and focus on post-machining of DED-built components. Simulation results suggest that the final residual stress states following machining DED structures are strongly tied to both the initial stress states and microstructure. Specifically, while pre-existing compressive or tensile bulk stresses have negligible effects on cutting forces, they play a significant role in shaping the final surface residual stress. In addition, refining the initial grain size is also found to increase the magnitude of compressive residual stress on post-machined surfaces. By revealing the mechanisms that drive the bulk residual stress formation in DED and the surface residual stress generation in machining, this thesis offers practical guidance for customizing both additive manufacturing and post-treatment technologies, thereby fostering the adoption of advanced manufacturing technologies in mission-critical industries.