Additive friction stir deposition (AFSD) is an emerging solid-state additive manufacturing process that leverages the friction stir principle to deposit porosity-free material. The unique flow of material that allows for the transformation of bar stock into a near-net shape part is driven by the non-linear heat generation mechanisms of plastic deformation and sliding frictional heat generation. The magnitude of these mechanisms, and hence the total applied thermal power, implicitly depend on the thermal state of the system, forcing power input to become a dependent variable. This is not the case in other 3D printing methods; thermal power can be controlled independently. In this work, the heat generation in AFSD is explored, and its transfer is quantified. In particular, the time-dependent ratio between the amount of conduction into the AFSD tool versus into the substrate is quantified. It was found for the conditions tested with a single-piece AFSD tool, conduction up the tool was on the order of the conduction into the stir. For a more modern three-piece tool, the ratio between the tool and the substrate varied between 0.3-0.1. It was also found that traversing faster resulted in more heat flux into the substrate as would be expected by moving heat source modeling. The total heat generated was also quantified as being equal to between 60% and 80% of the mechanical spindle power depending on the tool type and the exact process conditions. That ratio was found to be time-invariant. At the same time, this changing heat flux ratio was shown to dramatically alter thermocouple measurements in the tool, showing the uncertainty of that method of process control. The contact state between the stir and the tool was treated as a thin conductive layer and a contact heat transfer coefficient was calculated on the order of 20 frac{kW}{m^2K}. The limitations of this treatment were found to occur when a significant amount of the heat generation came from frictional heating rather than plastic deformation. This implies that any measurement conducted in the tool is related to the stir by a complex function driven by the state of the stir; showcasing the need for more well-understood in-situ monitoring. Finally, some of the ideas about thermal control are applied to a case study on the repair of corroded through holes using AFSD to restore fatigue life. It was found that modifying the thermal boundary conditions and applying active cooling at the end of the repair could improve the fatigue life drastically. This was due to less time spent in a thermally active region leading to less heterogeneous nucleation and less grain boundary nucleation. This more preferred microstructure morphology led to a change in the fracture mode and increased the number of cycles to crack initiation and the number of cycles after crack initiation. / Doctor of Philosophy / Metal 3D printing of industrially relevant aluminum alloys is plagued with problems. Additive friction stir deposition seems well posed to overcome some of the problems associated with aluminum printing. Being able to 3D print these alloys with properties that are as good as traditional manufacturing offers a large potential cost and time savings over traditional manufacturing for the aerospace industry (e.g. Boeing, Lockheed Martin, U.S. Navy). For these components to be part of a plane, the manufacturer must prove the components were made the same way print-to-print regardless of the actual shape of the component being made. This dissertation focuses on the key metallurgical variable of temperature and explores how thermal energy is generated and where that energy goes in to the system. The key takeaway is, that without precise knowledge of the total heat generated and the entire thermal system, assurances about processing temperature cannot be made. An exploration of heat generation and metrics about its dispersion are presented. This is followed by a study on repairing structural components while changing the thermal system to understand its effects.