Constrained Friction Processing (CFP) is a new friction-based technique suitable for the thermo-mechanical processing of lightweight metallic materials. The process has the potential to overcome challenges related to the processing of Mg alloys, related to their low ductility and workability at room temperature. During the CFP, the rotating shoulder is plunged into the plate, plasticizing the material to be extruded into the tool cavity created by the retraction of the rotating probe. The probe is a second rotating tool, which is placed inside the shoulder’s cavity and acts to constrain the plasticized material. The heat and shearing conditions developed during the process, responsible for the substantial microstructure refinement of the material, result in a complex material flow, which can be described by the combination of two shear components: simple shear and pure extrusion. Although a deep understanding of the material flow is a key aspect to enable the microstructural tailoring of materials and optimize their properties for specific applications, a comprehensive understanding has been, so far, not achieved. This project aims to provide a multi-scale description of the material flow during the CFP using an empirical modeling approach. It is expected that the models will provide extensive knowledge to support the comprehension of the rods’ microstructure, and thus, predict their properties. In the first part of the project, an empirical model for the material flow will be established to describe the shearing direction at each point of the rod. This approach will be founded on the findings that the shearing direction on the materials determines the texture of the rod. Rods will be produced and extensively characterized in terms of texture using electron backscattered diffraction, whose pole figures will allow the determination of the shearing direction. The collected data will allow the establishment of the model considering multiple independent variables, such as position along rod’s radius and height and rotation speed (RS), previously identified as the main parameter to influence the microstructure. The second part of the project proposes to investigate the effect of variations of RS to influence the macro material flow. For that, rods will be produced with additions of carbon nanotubes, which will act as tracer materials. The analysis of the macrostructures will disclose the material flow layers in terms of amount and distance between layers and allow the establishment of an empirical model to describe the variation of these features along with variations on RS. Additionally, Al wires will be also added to the process, and the analysis of the rods by 3D X-ray tomography will enable the estimation of strain over the CFP. The proposed models will be validated using scaled-up rods, and the interchangeability with other friction-based processes, exemplified here by friction extrusion, will be evaluated.